The discharge & the receiving water

Every water finding is a fraction. The numerator is what a facility puts in the river; the denominator is what the river has to dilute it with. Get the denominator wrong and every number downstream of it is wrong by the same factor — which is exactly what happened here.

There is no dilution

Table 12 of the fact sheet gives the design low flows at the outfall^rec : an annual 7Q10 of 0.21 cfs against the plant’s 23.208 cfs average design flow — roughly 110 times the river’s low flow. Ohio EPA does the arithmetic itself on page 13:

Acute Dilution Ratio = (1Q10 + WWTP flow rate) / WWTP flow rate = (0.17 + 23.2) / 23.2 = 1.0

A ratio of 1.0 is the floor of that calculation. At design flow, below River Mile 56.42, the Blanchard River is the effluent — and the acute wasteload allocation that follows is expressed as 30 percent mortality in 100 percent effluent. [verified]

The denominator that was wrong

This network used to screen Findlay against 8.67 cfs, a statistical fit at USGS gage 04189000, which put the effluent at 2.68× the low flow. That number is retired for this reach on two grounds, neither arguable: [verified]

The gage sits below the outfall — about 1.1 miles west-northwest, on a reach the Blanchard runs westward, so the flow it measured already contained the discharge it was being used to screen. And the gage is regulated: the station’s own REMARKS record that water is diverted upstream into Findlay Reservoir and released for low-flow augmentation. A gage downstream of a municipal augmentation scheme measures a managed flow, and using it as an assimilative denominator credits the discharger with water the city itself is releasing.

Where the Blanchard is genuinely unregulated, USGS publishes a 7Q10 of 0.03 cfs, and 0 cfs for its Eagle Creek tributary. The 0.21 cfs Ohio EPA computed at RM 56.42 is consistent with those; the 8.67 never was. Corrected, the screen reads 0.009:1 — about 41× tighter than the figure it replaced.

Interrogate the gage before using its number. Two questions settle most of it: is it upstream or downstream of the thing I am screening, and is the flow it measures regulated by anybody? Both answers were printed in the station’s own metadata all along.

The phosphorus finding that was retracted

The rest of this chapter exists because a finding turned out to be stale, and the way that was discovered is worth more than the finding was.

The work on Findlay was commissioned around a clean, damning sentence: the sewage plant runs over its phosphorus allocation every reported year, shielded from consequence by a general-permit bubble. That was a correct reading of the evidence available when it was written. The first thing this chapter has to say is that the evidence moved.

What the TMDL says

The Maumee Watershed Nutrient TMDL assigns Findlay WPCF a spring wasteload allocation of 3.2 metric tons^rec of total phosphorus, restated as a 3,200 kg Individual Load Limit by general permit OHP000001. Appendix 4’s record of what the plant discharged runs 2017 through 2021: 4.8, 5.5, 5.3, 5.5 and 5.4 metric tons — over the allocation by 50 to 72 percent in every reported year. [verified]

Read that table’s columns before its numbers. Table A4.5 prints seven numeric columns per facility and they are 2008 / WLA / 2017 / 2018 / 2019 / 2020 / 2021 — not a left-to-right time series. Findlay’s row is 4.4 · 3.2 · 4.8 · 5.5 · 5.3 · 5.5 · 5.4. Read naively, the first two numbers look like the start of a trend; they are a 2008 baseline and the allocation itself.

And on this record it was not a violation of anything. OHP000001 Part IV.C.3 makes an individual exceedance a violation only when the 39-facility cumulative load also breaches its cap. In the 2024 spring season the group reported 43,304 kg against a 64,170 kg cap. So long as the bubble holds, a plant running half again over its own allocation is compliant by the permit’s own terms. [verified]

Rebuilding the number

After 2021 the individual record stops — the 2024 Biennial Report publishes only the group total. It is still answerable, because the general permit publishes its own load equation (Load = C_M × Q_S × F) and every input is in the reported effluent record. Reconstructed from the DMRs: [inference]

The 2023 season — the last spring before the general permit took effect — is consistent with the TMDL’s 2017–2021 record and comfortably over. The 2024 season, the first the general permit governed, is 1.7 metric tons lower. The plant is not running at 5.4 MT any more, and the framing this site was commissioned on no longer describes it.

What changed is seasonal and visible in the monthly data: before the general permit the two halves of the year are indistinguishable; from 2024 onward the plant runs ~0.40 mg/L inside March–July and ~0.75 mg/L outside it, against a concentration limit of 1.0 mg/L that never changes. The non-spring concentration did not rise, so this reads as extra treatment during the season rather than load pushed out of it — though the corpus holds no operating record, so the cause is inference.

The trend 4,450 → 2,756 → 3,101 kg is not a trend toward safety, and the live question is no longer whether the bubble is holding but whether the margin is closing. The authoritative figure — parameter 51451, the permittee’s own calculated load — is not withheld so much as unreachable: it does not appear in this permit’s effluent chart at all, and pulling the general permit returns a chart with zero DMR rows. August 20 is the cheapest check on this page: re-pull the effluent chart after the July eDMR is due and look for 51451. If it appears it supersedes every derived figure above.