Open investigation · Van Wert, Ohio

Open leads

Every gap we're chasing on this site, in the open. Each lead is unverified inference until a source corroborates it — and every one traces to the corpus-completeness audit or a working hypothesis. Pick one up, answer it, or bring the document that closes it.

Submit a lead

open leads

withheld / sealed

under review

closed recently

20 open leads · unverified

Open question Unanswered [open] data/extracted/van-wert/incentive-water-instruments.yaml · data/extracted/van-wert/mega-site-instruments.yaml · data/reference/oepa/cooling-reconciliation.yaml

Three different fill volumes, two incompatible framings, and no instrument behind any of them

The City's Safety-Service Director announced ~660,000 gal as a ONE-TIME fill (2026-04-21); a County Commissioner described 700,000 gal that "will last 6 to 8 years"; the since-deleted project microsite carried 5,500 gal. The same ~660,000 gal is framed BOTH as a one-time fill and as an annual draw depending on who is speaking, and that ambiguity is the whole lead — at 660,000 gal/yr the draw is 0.0018 MGD, below the 0.01 MGD floor and consistent with a dry loop; as a one-time fill it says nothing about ongoing use at all. ⚠️ EVERY ONE OF THESE FIGURES IS THE CITY'S OR A CITIZEN'S, NOT QTS'S. The operator's own live FAQ declines to state a volume ("can vary widely … hard to predict"), so no number here is an operator disclosure and none of them can corroborate the closed-loop claim. Per the B2 reconciliation (#1682) the disclosed figure lands on `disclosed_makeup`, which by rule never feeds the classifier and never upgrades the source: a self-report of the very claim under test is not an instrument. The outcome stays a `gap` with the [reference] cooling pin KEPT.

Six days after his own press release the same director told Council the City "would be incapable" of supplying fill water. Both statements are in the committed minutes. #1409

Answer this → FILL-VOLUME

Open question Unanswered [open] data/documents/van-wert/operator/2026-08-05-qts-van-wert-project-faq.q-com.html · data/extracted/van-wert/incentive-water-instruments.md

The operator's FAQ names the one document that would settle campus water capacity

QTS's public FAQ (captured 2026-08-05) rests its water answer on "the analysis completed by QTS and City engineering". That document is not in the corpus and has never been public-noticed. It is the single highest-value records request on this site: it is the only named artifact that would replace the whole fill-volume dispute above with a figure, and it sits with a public office (City engineering), which makes it reachable under R.C. 149.43 in a way the operator's own internals are not.

The same FAQ says the loop "uses only water" while the City's sealed record says water AND glycol, and its Van Wert economy answer carries Richmond County (Augusta GA) boilerplate. #1409

Answer this → QTS-CITY-ENGINEERING-ANALYSIS

Signal Under review [inference] data/extracted/van-wert/water-watch.yaml · data/extracted/oepa/van-wert/2PD00006.fs.npdes.yaml

Van Wert drinks and discharges the same Town Creek flowline

Snapped to NHDPlus v2 via NLDI, the City's water plant (OHG8P0006 / OH0135569), its five CSO outfalls and WWTP outfall 001 (RM 13.87) all fall on ONE 17.05 km flowline, COMID 15653063, running south-to-north. The intake is UPSTREAM of the outfall, so nobody drinks the effluent — that is the [verified] part, and this lead does not dispute it. What is `[inference]` is the consequence: a campus that routes cleaning and blowdown water to the WWTP adds load to a plant already under a dated CSO construction schedule, on a creek with no summer or winter 7Q10 at all, upstream of the same city's own raw-water reservoirs. That is a different and more tractable question than the "will it discharge to Town Creek" the public debate was having.

Reservoirs hold 1.01 BG against a creek yielding 180 MG-1.26 BG/yr since 1951. #1409

Help confirm → SHARED-TOWN-CREEK-REACH

Open question Unanswered [open] data/extracted/van-wert/mega-site-instruments.yaml · data/documents/van-wert/council/26-05-028.pdf

The runway was laid across 2014-2016, and none of the three agreements that laid it is in the corpus

Ordinance 26-05-028 recites three Pre-Annexation Agreements between the City, township trustees and the Board of County Commissioners — Ridge 2014-11-14, Hoaglin 2014-12-22, Pleasant 2016-06-14. None is committed. They are priority 1 because they date the assembly a full decade before the 2026 announcement and are the only instruments that would show what the City committed to, and to whom, before any operator was named. Note what they also retire: the recurring "industrial since 2007" claim in press and public comment is supported by NOTHING in the committed set.

#1401

Answer this → PRE-ANNEXATION-AGREEMENTS

Redaction Withheld [open] data/extracted/van-wert/mega-site-instruments.yaml · data/documents/van-wert/council/26-05-028.pdf

No signed or certified copy of any of the six ordinances is public

All six ordinance PDFs the City published were uploaded FOUR DAYS BEFORE the vote and carry an unfilled "Passed this ___ day of ___". No signed or certified copy has been published. That is what makes the Type 1 / Type 2 contradiction unsettleable from the public record: the committee motion (4/27), the posted ordinance (5/07) and the agenda all say Expedited Type 2, the passage title says Type 1, and the minutes claim consistency "with the heading of Ordinance 26-05-028 itself" — a heading that reads Type 2 (verified visually at 300 and 450 DPI). A certified copy from the Clerk is the only thing that resolves it.

⚠️ vanwert.org/ordinances/ is a ROLLING WINDOW of pending items only — the May 2026 ordinances were already gone by 2026-08-03. The open WP REST media index enumerates the whole archive. #1401

Fill the gap → CERTIFIED-ORDINANCES

Open question Unanswered [open] data/extracted/van-wert/mega-site-instruments.yaml · data/reference/van-wert/parcel-assemblage.geojson

The hearing notice names four parcels; the assemblage has five

The published hearing notice names four parcels totalling 742.75 ac — 82.4% of the 901.698 ac the ordinances zoned. Parcel 17-034718.0200 (157.84 ac) is unnamed in it, while the committed auditor CAMA has it deeded to QTS VAN WERT LLC with the rest. The likeliest explanation is a post-petition split, which the county survey references would show; it is NOT established, and the alternative — that 17.5% of the campus was zoned without being named in its own hearing notice — is the reason this stays open rather than being filed as a clerical note.

#1401

Answer this → PARCEL-GAP-4-OF-5

Open question Low confidence [open] data/extracted/van-wert/bosc-site-footprint.yaml · data/extracted/van-wert/mega-site-instruments.yaml

1.1 acres between the zoned area and the deeded holding, unattributed

Three independent acreages now agree to within ~1.1 ac — 901.698 zoned (Ordinance 26-05-028 Exhibit A), 901.502 planar (the committed geometry in UTM 16N), 900.59 deeded (auditor CAMA). The residual is recorded and deliberately NOT attributed to right-of-way or to survey method. Logged at `low` because it is now a rounding-scale question rather than a finding: it was carried as a 61.4-ac gap until #1401 read the annexed area off the instrument and retired the press's "~962 ac", and the honest thing is to keep the residual visible at its real size rather than close it.

#1403

Answer this → ANNEXATION-RESIDUAL-1-1-AC

Open question Unanswered [open] data/reference/van-wert/parcel-assemblage.geojson · data/extracted/van-wert/bosc-site-footprint.yaml

The Marsh Foundation → Thor → QTS deed chain is not closable from the parcel layer

The county's AGOL CAMA layer carries no owner mailing address, no legal description, no grantor and no deed book/page, so the conveyance chain cannot be closed from committed geometry — it is a Recorder pull. The anchor parcel's own price step is on the record and unexplained by it: 221.15 ac sold 2025-08-22 for $10,394,000 (~$47,000/ac) and again in June 2026 for $110,575,000 — exactly $500,000 × its CAMA acreage, a 10.6× step in ten months. ⚠️ The four same-day parcels carry ONE date and ONE consideration across all four, the signature of a single multi-parcel deed, so that amount is recorded verbatim per parcel and must NOT be summed; the campus's total consideration stays open.

Route: Van Wert County Recorder at countyfusion14.govos.com, grantor "Marsh Foundation" — a session login, not scriptable without one. #1404

Answer this → DEED-CHAIN-LLC-GRAPH

Open question Unanswered [open] data/extracted/oepa/van-wert/2GC08872.noi.npdes.yaml · data/extracted/van-wert/regulatory-watch.yaml

The air permit-to-install is a dated instrument-grade negative, and it is what would move the facility domain

The campus's construction-stormwater NOI certifies, under penalty of law on 2026-07-21, that its Ohio EPA air permit-to-install status is YET_TO_APPLY. That converts "no PTI found" from an unsuccessful search into a dated negative on the applicant's own word — and it is the single instrument that would legitimately lift `facility` from `seeded` to `live`, since #1630 grades the domain on documentary depth and Van Wert's load is an announced [reference] ceiling while its cooling is an operator FAQ claim. Re-swept 2026-08-05: still dry (35 QTS documents in the eDoc portal, none newer than 2026-07-30). QTS's whole Ohio air fleet is Licking County. ⚠️ Do NOT reach for `live` by re-grading the load or the cooling. The network's rule is to let the domain lock and ask for the source.

The eDoc sweep sorts date-DESC, so page 1's oldest date bounds the whole result set — that is what makes the zero a controlled zero rather than an empty query. #1408

Answer this → AIR-PTI-YET-TO-APPLY

Open question Unanswered [open] data/extracted/van-wert/water-watch.yaml · data/extracted/van-wert/regulatory-watch.yaml

The *WD modification's first CSO construction milestone came due 2026-08-01 and is untested

NPDES 2PD00006*WD put TWO milestones on 2026-08-01 — event 90199 (annual report) and event 3099 (begin CSO construction). The 2026-07-28 filing completed only 90199. Event 3099 is open and unverified. It matters beyond compliance bookkeeping: three permit-schedule events (CSO Event Report, O&M Report, Combined Sewer Report) have been "unachieved and not reported" continuously since early 2024 with ZERO formal actions and $0 penalties, so the driver of this permit's 12-of-13 non-compliant quarters is reporting, not effluent.

next check 2026-08-17 #1408

Answer this → CSO-EVENT-3099

Open question Unanswered [open] data/extracted/van-wert/water-watch.yaml

The instrument holding the CSO overflow total is the same one that is not being filed

A monthly CSO volume TOTAL is not derivable from the DMR: it carries a monthly average and a daily maximum and no total, and monthly_avg × occurrences fails against the record (Wall St., May 2026 — 1 occurrence, 0.142 Mgal average, 0.277 Mgal daily maximum). Only a FLOOR — the sum of monthly daily maxima — is defensible. The instrument that holds the actual total is the CSO Event Report, which is one of the three schedule obligations unfiled since 2024. The gap and the reporting violation are the same object, which is what makes this worth stating rather than treating as a data limitation.

⚠️ ECHO's standardized value for parameter 74063 is wrong by 1e12 (it divides where it should multiply) — read LimitUnitDesc, never StdUnitDesc. Confirmed on two permits in two counties. #1408

Answer this → CSO-EVENT-REPORT-VOLUME

Open question Unanswered [open] data/extracted/oepa/van-wert/2GC08872.approval.npdes.yaml

No co-permittee NOI has been filed, so the general contractor is unnamed

Ohio EPA's approval letter expressly notes that each additional operator engaged at the site must file its own Co-Permittee NOI. None is in the corpus as of 2026-08-05, so the general contractor is open. This is a live watch rather than a settled absence: Sidney's campus produced exactly such a co-permittee filing, so the form does get used in this state and its absence here is a fact about this project's stage.

#1408

Answer this → CO-PERMITTEE-NOI

Open question Withheld [open] data/extracted/oepa/van-wert/2GC08872.noi.npdes.yaml

A nationwide-permit verification is never public-noticed, so silence here can never become absence

The NOI answers NO to an individual 401 Water Quality Certification and NO to an Isolated Wetlands Permit while answering YET_TO_APPLY on the USACE nationwide permit. Those are the applicant's assertions on a form, not agency determinations, and no 401 or isolated-wetland instrument for this campus is in the corpus. ⚠️ Recorded as `withheld` rather than `unanswered` on purpose: an NWP verification is NOT public-noticed, so no amount of watching the state portal will ever turn this silence into a documented absence. Only a USACE Buffalo District (LRB) FOIA answers it. Do not log a dated negative against a route that structurally cannot produce one.

#1408

Answer this → USACE-NWP-VERIFICATION

Open question Unanswered [open] data/extracted/grid/van-wert/van-wert-haviland-138kv.project.yaml · data/extracted/van-wert/mega-site-instruments.yaml

No document in the corpus states this campus's MW, including the transmission case named for it

The 500 MW carried on the profile is a [reference] bracket from the developer, never an operator disclosure — QTS's own FAQ declines to state capacity, and AEP Ohio told Council it had ">30,000 MW" of early interest against 5,342 MW under signed agreements while stating NO MW for this campus. OPSB case 25-0697-EL-BLN (approved 2025-11-21, unbuilt) names the Van Wert Mega Site in its Statement of Need and counts 30 requests for transmission service in the past year — but "QTS" and "data center" appear on NONE of its 612 pages. Top pull: AEP Ohio's 2025 Long-Term Forecast Report, pp. 49 and 76-77 (the filing's own Appendix B) is the most likely public document to carry a Van Wert load figure.

⚠️ PJM's interconnection queue is GENERATION-only — large load never appears in it, so do not watch the queue. Load reaches PJM through the transmission owner. #1408

Answer this → AEP-LOAD-FIGURE

Signal Under review [inference] data/extracted/grid/van-wert/van-wert-haviland-138kv.project.yaml · data/extracted/van-wert/incentive-water-instruments.yaml

The City disclaims the Haviland 138 kV line as data-center-related, and the record half-agrees

The Safety-Service Director told Council on 2026-05-27 that AEP had planned the Haviland line "for about 3 years" and that "this project was not related to the Data Center". That cuts against attributing OPSB 25-0697-EL-BLN or its $45,877,232 cost to this campus — while the filing's own Statement of Need does name the Mega Site. Both are on the record and this lead does not resolve them; it records that the attribution is contested by a named official. Two details that keep the reading honest: the project is an existing 69 kV line rebuilt as a double circuit DESIGNED at 138 kV but INITIALLY OPERATED AT 69 kV — not a 138 kV line on day one — and the approved minutes read "135-kilovolt", which matches neither figure.

Cost is recovered via the FERC formula rate allocated to the AEP Zone, not to this campus. #1408

Help confirm → HAVILAND-LINE-ATTRIBUTION

Open question Unanswered [open] data/extracted/van-wert/incentive-water-instruments.yaml · data/documents/van-wert/council/7.13.26.pdf

An incentive package exists as a negotiation and as no instrument

Five named routes were searched to a dated negative on 2026-08-05: the City's own legislation and minutes through Ord. 26-07-040, the County Commissioners' published 2026 agenda index (63 meeting dates), the Lincolnview Local BOE via BoardDocs, the Ohio Tax Credit Authority's 2026 minutes, and the committed ODD statewide tax-incentive export. No enterprise zone, CRA, PILOT, TIF, JEDD or development agreement exists for this project in any of them. What DATES the negative is the City's own record: the 2026-07-13 Economic Development committee report says VWAED's director "is concluding incentives and development agreements for the data center and hopes to have those done by the next council meeting". One does not conclude what is already signed — so as of that date the package was live and unexecuted.

Re-check the 2026-07-27 and 2026-08-10 minutes; neither was published as of 2026-08-05. #1407

Answer this → DEVELOPMENT-AGREEMENT

Open question Unanswered [open] data/extracted/van-wert/incentive-water-instruments.yaml

The state data-center sales-tax exemption is paused, and this project has never requested one

The Governor directed the Tax Credit Authority chair on 2026-05-27 to pause data-center exemptions — two days before the joint City/QTS announcement. The first meeting wholly under the pause (2026-06-29) took seven new projects and zero data-center exemptions. No Van Wert or QTS item appears in any 2026 TCA meeting read. ⚠️ Do NOT read "Van Wert Forward II" on the 2026-06-29 TCA agenda as a Mega Site award — it is a 2022 historic-preservation/TMUD downtown project. And do not read the Cologix award of 2026-06-01 as the pause failing: the pause covers NEW requests, and that one predates it.

The committed ODD export lags — its newest Datacenter Tax Exemption is 2022-08-29 and it does not carry the 2026-06-01 Cologix award its own TCA minutes record. #1407

Answer this → STATE-EXEMPTION-STATUS

Signal Under review [open] data/extracted/van-wert/regulatory-watch.yaml

The local ballot route against the zoning is closed, and it was an initiative rather than a referendum

A petition filed 2026-07-18 against Ordinance 26-05-029 was certified INSUFFICIENT on 2026-07-27 — 255 valid signatures against 323 required. There is no cure period. The route is foreclosed. ⚠️ It was an INITIATIVE, not a referendum, and the distinction is not pedantry: a referendum suspends the ordinance pending the vote, an initiative does not, so the zoning was never stayed at any point. Recording it as a failed referendum would misstate what was legally at stake.

#1408

Help confirm → REFERENDUM-FORECLOSED

Open question Under review [open] data/extracted/van-wert/incentive-water-instruments.yaml · data/documents/van-wert/council/26-06-034-DRAFT.pdf

Ordinance 26-06-034 does not do what Council moved to do

Council moved to PROHIBIT new private wells. The drafted Exhibit A makes new wells a CONDITIONAL USE in every zoning class — no prohibition, no grandfather clause, and no geothermal exception in its own text. The gap between the motion and the instrument is the lead; which one governs depends on a signed and certified copy, which for this ordinance is also not public.

#1407

Answer this → PRIVATE-WELL-ORDINANCE

Claim Low confidence [inference] data/extracted/van-wert/mega-site-instruments.yaml

Three benefit figures circulate publicly and appear in no instrument

"$200M over 20 years", "200-250 jobs" and "$80,000" (an average wage) appear ONLY in the written public-comment campaign and in press repeating it — never in an ordinance, agreement or filing. They are recorded here as unattributed claims so that a later instrument can be tested against them, not because the corpus gives them any standing. A second reading of the same file supports the caution: 21 of the 88 written-comment pages repeat an earlier body verbatim (67 distinct bodies; one submitted five times), so the volume of comment is not the volume of independent comment.

#1401

Corroborate → LETTER-CAMPAIGN-FIGURES

Showing all 20 open leads. Closed leads move into the record →