Nothing is built, nothing is owned, and nothing is left behind.
That sentence is either the most attractive or the most alarming thing about in-water remediation, depending entirely on which appraisal template an institution reaches for first.
Water finance is built around structures. Plants, networks, reservoirs, interceptors, treatment works: assets with a book value, a design life and a decommissioning obligation. That template has moved an enormous amount of capital, and it is the correct instrument for the source of nutrient pollution.
It is the wrong instrument for the condition of a receiving water body during the decades that source work takes. Applied there it produces a familiar outcome: a proposal appraised against criteria it was never designed to satisfy, failing on residual value and asset security rather than on whether it delivers anything.
So it is worth putting the two side by side properly.
Comparison
Two instruments, both legitimate, doing different jobs.
| Dimension | Catchment capital works | In-water remediation |
|---|---|---|
| What is acquired | Structures, land, easements, network | Contracted capacity in a defined volume of water for a defined window |
| Time to first benefit | Years to decades, after construction and behavioural change upstream | Within a season, measurable on instrumentation during the same window |
| What it solves | The source of nutrient and sediment loading — permanently, if sustained | The condition of the receiving water — temporarily, and only while contracted |
| Reversibility | Low. Built infrastructure commits a corridor and a budget for decades | High. Capacity can be scaled, relocated or stopped between seasons |
| End of term | Residual value, handback condition, decommissioning liability | No residual asset, no handback, no decommissioning — and no security either |
| Underwriting basis | Physical asset and regulated revenue | Contracted service and measured outcome, which requires verification design |
| Regulatory pathway | Planning consent, construction and abstraction permits, then upstream compliance held for decades | A discharge consent against a stated limit at the outlet — the instrument a ministry writes routinely, not one it has to invent |
| Principal risk | Construction, cost overrun, upstream compliance over decades | Efficacy at scale in open water: no result has been published, by this operator or by anyone competing with it |
The last row is the one an investment committee will spend its time on. The row above it is the one that moved. Because the treatment is contained — water drawn into the hull, processed, released against a number — the approval being sought is a discharge consent rather than authority to put an oxidant into a coastline, and the second of those has no settled framework in most jurisdictions. Approval risk and schedule risk are the same risk in this sector. The evidence behind both rows is set out below.
Precedent
Water services have industrialised before, and quickly.
An operating water service is usually appraised as permanently expensive. The recent history of this sector says otherwise.
The World Bank reports that the cost of desalinated water has fallen from around US$5.00 per cubic metre in the 1980s to as little as US$0.40 to US$0.50 in recent projects. That is roughly an order of magnitude, driven by membrane development, energy recovery, scale and operating discipline rather than by any single breakthrough.
Operating discipline alone can move remarkable amounts. Dubai's water authority cut non-revenue water losses from 42 per cent in 1988 to 4.6 per cent by 2023 — a change in what a utility manages, not in what it builds.
Neither precedent proves anything about in-water remediation costs. What they establish is that the appraisal instinct to treat a novel water service as permanently expensive has been wrong before, in this exact sector, within the professional memory of the people doing the appraising.
Demand
The addressable problem is growing on a documented curve.
Diaz and Rosenberg's 2008 study in Science counted more than 400 coastal hypoxic zones worldwide, covering roughly 95,000 square miles, and found the count had approximately doubled every decade since the 1960s. Few environmental trends are that clean, and none of them are encouraging.
The Baltic is the instructive case for anyone modelling how long a regional recovery takes. Across 1961 to 2005 its hypoxic zone averaged something like 50,000 square kilometres of bottom water below 2 millilitres of oxygen per litre, with substantial seasonal and inter-annual variation. That is a semi-enclosed sea with nine states on its shores, decades of monitoring and one of the longest-running coordinated nutrient efforts anywhere.
An institution planning across decades should read those two facts together. Demand for this capability is not speculative, and the alternative pathway is slow even where governance is strong and money is available.
Sources: Diaz and Rosenberg, Science, 2008 · Conley and colleagues, Environmental Science and Technology.
Three things outside the boundary
Boundaries first, because an allocation made on an inflated scope produces a write-down and a reputational problem at the same time.
Ocean acidification is not addressable by any vessel programme. NOAA reports aragonite saturation state falling by about a fifth since the pre-industrial era across much of the ocean surface, and its own dataset puts the working numbers plainly: above a saturation state of 3, calcifying organisms are more likely to survive and reproduce; below 1, aragonite shells begin to dissolve. Local pH support inside a bounded nursery or hatchery intake is real and defensible. Reversing a basin-wide trend is not on offer from anybody.
Basin-wide marine heatwaves are the same category. Bleaching-level heat stress affected 21 per cent of reefs in the first global bleaching event of 1998, 37 per cent in the second, 68 per cent in the third and 84 per cent in the fourth. In-water cooling addresses local hot spots. It does not address the trend those figures describe.
And nutrient loading is generated on land. It will be reduced on land, by agriculture and wastewater capital, or it will not be reduced. A national programme that funds in-water work while abandoning catchment reform has bought itself time and then wasted it.
The evidence, document by document
Four categories, each with the body that produced it named alongside.
What is established
Nanobubbles are a characterised class of object rather than a marketing term, and the gas-transfer consequence has been measured in peer-reviewed work. None of it is proprietary to this group and none of it is contested.
What is validated, and by whom
Two NOAA National Centres for Coastal Ocean Science validations exist in this space: a freshwater algal remediation trial at pond scale in 2018, and a ballast-water invasive-species evaluation in 2020. Both name other companies' technology and other partners. They support the approach and its safety at those scales. They do not validate any particular vessel, and any document presenting them as company-specific validation is misrepresenting them.
The 2020 release carries an abbreviation worth expanding in full. The system it describes is another party's Nanobubble Ozone Technology, which shortens to NBOT — the same four letters this network uses for Nano Bubble Oxidation Technology. Two systems, two owners, one abbreviation, and no relationship between them.
What only the operator can attest
Alarivean holds a Cooperative Research and Development Agreement with NOAA. Agreements of that kind are private instruments and are not published, so there is no filing to pull and no register to search. Under it, work with the NOAA scientist Dr Peter Moeller found the treatment switching off the functional groups that make a toxin toxic, rather than settling the material out of solution. That is a research relationship with a federal science agency, stated on the operator's own authority. It confers no approval on anything sold in this group, and a data room should treat it as what it is.
What is unresolved in the science
Whether nanobubbles themselves generate hydroxyl radicals is genuinely disputed: a 2020 study from Moleaer and Arizona State University reported reactive oxygen species from injected nanobubbles, and a controlled 2023 study by Chae, Kim, Kim and Fortner in ACS ES&T Engineering found generation minimal, if it occurred at all, under the conditions tested. Programmes are built on the mass-transfer result and the engineered oxidation stage, both measurable. An allocation here is underwritten against those two and never against radical chemistry.
What the field record is
In September 2024 Alarivean, Inc. of Scottsdale, Arizona received Florida Department of Environmental Protection permit FLOA00062, issued under Chapter 403 of the Florida Statutes with a five-year window, to field-test an ozonated-seawater craft for red tide mitigation, conditioned on active Karenia brevis above 100,000 cells per litre. The Sarasota Bay Estuary Programme and the Florida nonprofit START both documented it independently. What the estuary programme has said publicly is that open water remains unanswered in the published record. That is correct, and it is also why the programme wants the work.
On Alarivean's account, the estuary programme backed that application with technical input and, deliberately, with none of its own funding. A body that has paid for nothing is free to publish whatever the water turns out to show. What it wants is the measuring role in open water, so that any endorsement reaching past the bay is built on data its own people gathered, and Alarivean wants it holding that role. The next evidence step follows from it: instrumented open-water work, with the local science body reading the results.
Meanwhile a permit to test is not evidence of scale, and an institution should structure accordingly: assessment, then bounded calibration with pre-agreed viability parameters, then contracted capacity, with a genuine option to stop after each.
Where each of these was published
- World Bank — From scarcity to sustainability: the GCC's journey towards water security, March 2024.
- Diaz and Rosenberg, Science, 2008 — global count of coastal dead zones.
- Conley and colleagues — Baltic Sea hypoxia extent, Environmental Science and Technology, 2011.
- NOAA Ocean Acidification Program — acidification is more pervasive than previously thought, for the fifth decline in aragonite saturation; and NOAA Science On a Sphere — the saturation-state dataset, for the values of 3 and 1.
- NOAA Coral Reef Watch and ICRI — the global bleaching event comparison across all four events, and the bleaching alert levels keyed to Degree Heating Weeks.
- Chae, Kim, Kim and Fortner — reactive oxygen species generation from nanobubbles, ACS ES&T Engineering, 2023.
- NOAA NCCOS — freshwater HAB validation, 2018 and ballast-water validation, 2020.
- Sarasota Bay Estuary Programme and START — public documentation of FDEP permit FLOA00062, 2024. Operating company material at alarivean.com.
Oversight as an underwriting term
The record has a custodian who is not us.
Every programme is built with continuous independent oversight inside it, and on an instrument underwritten against measured outcome that is not a governance nicety. It is the security.
Alarivean insists that relevant local scientific bodies are engaged for the life of a programme, with access to every water-quality feed the operator works from and to any others those bodies bring themselves. Human and environmental safety sit ahead of throughput in the risk controls, and somebody outside the company has to be positioned to say whether that held on the day.
For a committee, this is the line between a fundable service and a services contract taken on trust.
- Continuous, not bookended Not a consultation at mobilisation and a report at handover. The engagement runs through every season the vessels are in the water, including the quiet ones.
- Live feeds, not a digest The oversight party reads the same instrumentation we read, at the same moment we read it. A quarterly summary compiled by the supplier is a different product wearing a similar name.
- Local, because local is answerable A university department, a fisheries institute, an estuary programme — bodies that already carry a reputation in that jurisdiction and will still be carrying it in year six.
- Custody, which is where these instruments usually fail Outcome-based structures rarely come apart on the measurement. They come apart on who kept it. From the first treatment pass the record sits in more than one pair of hands.
- Required, not permitted The arrangement is a condition of the programme design and it survives a client who would prefer a quieter season. An investment committee can price against a record with two custodians. It cannot price against one.
From an investment committee
Six questions, answered without hedging
Where does this sit relative to catchment investment?
Alongside it, and second to it in permanence. Catchment work is the only durable answer to nutrient loading and should be funded on its own merits.
In-water remediation holds the condition of the receiving water across the fifteen to thirty years that work takes. Presented as a substitute, it is a false prospectus and should be refused.
What is the residual value at the end of a concession?
None, by design. No structure, no land, no plant to hand back or decommission, and therefore no stranded-asset risk and no residual-value negotiation.
The corollary is real: there is no physical security to underwrite against. The instrument has to be underwritten against a contracted service and a measured outcome, which puts unusual weight on verification design.
How is performance verified, and by whom?
In-situ profiles through depth at the point of treatment — dissolved oxygen, temperature, salinity, pH, turbidity — before and after each pass, with earth observation covering the wider zone.
Some of the standards already exist in public. NOAA Coral Reef Watch publishes an alert scale keyed to degree heating weeks: reef-wide bleaching risk from 4, mortality among heat-sensitive corals from 8, multi-species mortality from 12, and two further levels above that at 16 and 20. Verification should sit with a party whose fee does not depend on the reading.
What is the residual environmental liability of the treatment itself?
A discharge standard applied at the outlet, with no allowance made for the receiving water. Ozonating seawater can form bromate, and bromate is stable — it cannot be quenched and it cannot be diluted into compliance — so it is prevented before it forms rather than detected afterwards.
The water released is held at or below 10 micrograms per litre, a threshold drawn from drinking-water practice, and a batch that misses it is held aboard and landed instead of released. Inline instruments govern the treatment and can cut the oxidant on their own, but a stable end-product is invisible to them, so the confirming assay is an independent laboratory analysis. Both are needed. Neither covers for the other.
Which entity carries the liability, and under which law?
Alarivean, Inc., in the jurisdiction where the work happens. It signs the service contract, holds the environmental permit issued by your own regulator, employs the crew and carries the insurance.
Credible Ventures holds the intellectual property and no service obligation to a buyer, which is why the two sit in separate entities. A parent guarantee is a separate instrument: it has to be negotiated, written and signed by a named entity to exist at all.
Who is watching the operator while the programme runs?
Local scientific institutions, continuously, reading the same live feeds we read and any others they choose to add. Alarivean requires this rather than allowing it, and it is written into the design of a programme rather than conceded during one.
The consequence for an allocator is narrow and worth a lot: the performance record is not held solely by the party being paid on it.
Structured staging
Fund the assessment before funding the programme.
The coastline, the exposed sectors, the horizon and the counterparty structure you would use. Alarivean returns a staged assessment with the uncertainty attached rather than removed.