The treatable parts of a river are the still ones.
A free-flowing reach has replaced its entire volume before the crew has finished the log. Everything an operator can offer on a river follows from that one sentence.
What a river does to a result
A lake holds a result. A river exports one.
Treat water drawn from a lake and the benefit stays in the basin you paid for. Do the same in a bay and the tide takes a share of it and brings some back. Do it in a free-flowing reach and the oxygen-enriched effluent is gone, in one direction, at the speed of the current — while the water arriving behind it has never been near a hull.
Rivers are also good at fixing themselves in the one respect we sell. Turbulence over a bed, a weir crest, a rapid — every one of those is an aeration device, and a healthy reach re-oxygenates continuously without anyone spending a pound on it. That is why a river with a persistent oxygen deficit is nearly always a river where the flow has been taken away from it.
Impoundment behind a weir or dam. Canalization. A dredged navigation channel with a deep, still bottom layer. A tidal reach at neap tides in a dry summer. A marina cut into the bank with one opening. In each of those, water that used to tumble now sits, stratifies and behaves like a small lake with a current running over its lid.
Those are the places worth treating, and a river program is a list of them rather than a length of channel. It is also a standing list, worked round the year on subscription, because a still pool refills with exactly the same problem the moment it stops being worked. Where the problem is a hundred kilometers of open river, we are the wrong instrument, and fleet size changes nothing about that.
Where a hull earns its place
Five river settings, and what makes each one treatable.
The common factor is residence time. If the water in question stays where it is for days rather than minutes, there is something to work with. If it does not, there is not.
| Setting | Why the water stays | What usually fails there |
|---|---|---|
| Impounded reach | A weir or dam converts a flowing reach into a shallow reservoir with a slow bottom layer | Summer stratification, low oxygen at depth, blooms in the upstream end |
| Marina or harbor basin | One opening, no through flow, and boat traffic that stirs the surface but not the bed | Odor, surface films, fouling, complaints from people who moor there |
| Navigation channel | Dredged depth well below the natural bed, holding a dense, still layer | Oxygen deficit under the working depth; spill and discharge residence |
| Tidal reach at low flow | Slack water twice a day, weakly flushed on neaps and in drought | The classic estuarine oxygen sag, worst in a hot dry August |
| Intake approach | A pocket of water drawn slowly toward a plant screen | Bloom material and organic load arriving at a plant that has to keep running |
A calibration phase on a river is scoped inside one of these, on one consent, with agreed parameters — not across a catchment.
The number an alarm gets set from
The Delaware review's conclusion, and the appendix line people quote.
This matters more on rivers and estuaries than anywhere else, because river basin authorities are where dissolved oxygen standards actually get written.
The Delaware River Basin Commission commissioned a literature review of dissolved oxygen requirements for sensitive species in its estuary. Its conclusion: lethal-level requirements run between 4 and 6.4 mg/L, with roughly 5 mg/L adequate for many species and 6 to 7 mg/L protective.
A much lower figure from the same document circulates widely — somewhere between 0.5 and 1.0 mg/L, offered as though it were a general lethal threshold. It is not. It is a nineteen-hour result for juvenile White Perch alone, sitting in an appendix table, and an operator who set a night alarm from it would be an order of magnitude low. By the time that alarm sounded, the reach would have been uninhabitable for most of a season.
Errors of that kind survive being copied between documents for years, which is why a program takes its threshold from the review's conclusion and then from the species and season in the reach itself. What the number is then used for — alarm levels, stop authority, who may halt a pass — sits on the controls page.
Where the load ends up
The worst thing a river does happens several hundred kilometers after it stops being a river.
The Mississippi and Atchafalaya deliver roughly 2.76 billion pounds of nitrogen and 310 million pounds of phosphorus to the Gulf of Mexico, with agricultural sources contributing more than seventy percent of both. None of that damages the river visibly enough to make the evening news. It assembles itself into a hypoxic zone off Louisiana every summer instead.
The five-year average of that zone is 4,755 square miles, against a task-force target of under 1,900 by 2035. Two and a half times the goal, on a river system where the reduction programs have been running for decades.
So a river conversation is nearly always a basin conversation. The land-use answer is generational, unevenly funded and politically slow, which is why coastal states end up paying for a condition in a bay against a problem created eight hundred kilometers inland. Where the load enters through a countable set of hotspots rather than the whole floodplain, a firewall on those inflows keeps the dynamic from spreading while the slow answer is argued about.
HAB Control sets out how a basin strategy and an in-water intervention are written into the same document without either overstating what it can do.
Consent, before anything floats
A lake has an owner. A river has a committee, and sometimes two countries.
Abstraction sits with one authority and discharge with another. Navigation belongs to a third, flood management to a fourth, fisheries to a fifth, and the bed of the river itself may belong to a landowner who has never met any of them. Add a border and the whole set doubles.
What we do about that is limited and worth stating exactly. We do not run your permitting and we do not represent you to a regulator. We bring the method statement, the discharge and monitoring plan, the crew accreditations and an oversight arrangement with a local scientific institution, and we scope the calibration phase inside a single consent boundary so that one office can say yes without polling four others.
The rest is yours, and it is usually the longer half. A river program that looks slow is nearly always waiting on consent rather than on equipment, and that timetable belongs to the authorities holding it rather than to any contractor.
Where a spill or a discharge event is the trigger rather than a season, the duty changes shape and so does the paperwork — ChemSlayer covers that case, including why the capacity has to be contracted before the phone rings.
Straight answers
Four we get asked about rivers
Can you treat a whole river?
No. Treated water returned to a flowing reach is kilometers downstream inside the hour, and the water arriving behind it has never been near a hull.
River work means the places where flow slows or stops. Behind a weir, inside a basin, in a tidal reach at slack water, in the pocket in front of an intake. Those are treatable. A hundred kilometers of open channel is not.
Our problem is agricultural runoff. Can you help?
Yes, but not by changing what leaves the field. Agricultural sources contribute more than seventy percent of the nitrogen and phosphorus arriving at the Gulf from the Mississippi basin. Reducing that is land-use policy, and it runs on a twenty-year clock.
Where the loading enters at a hotspot rather than seeping off everywhere at once, the oxidation and aeration firewall can be established on that inflow to mitigate the dynamic before it spreads outward. Dr Peter Moeller of NOAA observed an effect on nitrogen and phosphorus; the work continues and no number goes on a web page until it is understood. Alongside that, in-water capacity holds a condition in a defined stretch. Both are promises we can put parameters against.
One thing this is not: an argument against fertilizer. Fertilizer feeds people. Optimize how it is used, cut the harmful pesticides, and handle the runoff where it arrives — throttling how food is grown to save a bay moves the harm rather than removing it.
What dissolved oxygen target would you work to?
Whatever your regulator sets. Where a site carries no standard, we start from the Delaware River Basin Commission's review, which puts the lethal-level requirement for sensitive estuarine species between 4 and 6.4 mg/L.
That band is then narrowed against the species and the season in your reach, and written into the stop conditions before a hull moves. The 0.5 to 1.0 mg/L figure circulating from the same document is an appendix result for one juvenile species, and no alarm is set from it.
What happens to the treated water when it leaves the reach?
First it goes back into the reach it came out of, at the same place, and then the current takes it downstream to somebody else — which is exactly why the limits get settled before anything is deployed rather than argued about afterward. The reach is held to figures your own authority sets for that water and the service level agreement carries. Nothing is impounded on deck waiting to be signed off, and there is no tank on the deck that could impound it: flow through the reactor is continuous, the instruments govern it in real time and can cut the oxidant unprompted, your oversight body reads the same feeds, and you can halt an active deployment on any suspected breach. A high-risk duty runs further back from the line, not into a wait.
On a river the evidence flows past within the hour, so the record has to be made in the water while the pass is happening. Reconstructing it afterward is not possible for anybody.
Name the reach
Which stretch, and who consents it?
Tell us the impoundment, the basin or the tidal reach, roughly how long the water sits there, what fails and in which months, and which authority would have to agree to anything happening in it. That last one decides more of the answer than the chemistry does.