What is settled, what is not, and what we refuse to claim
This page exists because the category is full of numbers with no method behind them. Separating the three kinds of claim is the fastest way to read any nanobubble brochure — ours included.
Three kinds of claim, three different burdens of proof
| Type | Example | What would prove it |
|---|---|---|
| Definitional | "This is an ultrafine bubble of 100 nm" | A named method and a printed distribution. Settled by ISO 20480-1/2. |
| Mechanistic | "Bubbles attach to floc and lift it" | Peer-reviewed mechanism plus a control experiment. Mostly settled. |
| Performance | "Raises yield 30%", "cuts energy in half" | A controlled comparison on your water, with the method stated. Almost never provided. |
"+30%" and "efficiency doubled" claims usually belong to the third row and arrive with no control, no stated measurement method, and no description of the water they were measured in. Without a control you cannot separate the bubble's contribution from the pump's and the mixing's — and the pump is much cheaper.
What ISO actually settled
ISO/TC 281 "Fine bubble technology" was formed in 2013, with its secretariat in Japan. It has published the 20480 series: Part 1 terminology (2017), Part 2 attribute classification (2018), Part 3 generation methods (2021). That is a real achievement, and it is also the limit of it: the series standardises names, classification and how to make bubbles. It does not standardise how much oxygen a generator must transfer, or what removal rate a flotation unit must reach.
The most common misuse of a standard number
"Certified to ISO 20480-1" is a category error. Part 1 defines vocabulary. Nobody certifies a machine against a terminology standard. If a tender document asks for it, the person who wrote the document has not read it — and that tells you something about the rest of the specification.
Four questions the literature has not closed
- Why do they persist? Epstein–Plesset theory says a 100 nm bubble in saturated water should dissolve in microseconds to milliseconds; internal pressure is higher still at 60 nm. Observation says hours, with reports beyond nine months in sealed glass. Charge-stabilisation, dynamic-equilibrium, high-density and contaminant-shell models all explain part of it; none explains all of it.
- Are the counted objects actually bubbles? A 2026 review of food-system measurements states it plainly: light scattering cannot prove a particle is gas. NTA reports "scatterers in this size band", not "bubbles". Some fraction of any count may be nano-droplets, aggregates or debris.
- Is there a dose–response relationship? For most applications nobody has published the curve of effect versus concentration. Without it, "more bubbles" is an assumption, not a design rule.
- Do two instruments agree? NTA, DLS, laser diffraction and resonant mass measurement give different answers on the same sample. Cross-laboratory comparability is weak, so a number is only meaningful with its method attached.
Six situations where we advise against buying
- You cannot state the target parameter. "Improve water quality" is not a target. If you cannot name the number that must move, no supplier can be held to it.
- There is no baseline measurement. Without three months of pre-installation data, any post-installation improvement will be attributed to the equipment whether or not it caused it.
- The problem is hydraulic, not chemical. If the basin is short-circuiting or the diffusers are blocked, fix that first — it is cheaper and it will work.
- The water is at 70 °C and saturated with carbonate hardness. Bubble lifetime collapses and the descaling mechanism has no data behind it.
- The scale is silica or calcium sulphate. Published evidence covers calcium carbonate systems. For silica and gypsum we have no defensible claim, and we will say so.
- You need the result in two weeks. Biological and fouling systems respond over months. A supplier promising a two-week step change is promising a measurement artefact.
Four controls that make a trial meaningful
- Run a parallel control. Same water, same flow, same duration, no bubbles. Two basins or two identical loops, not before-and-after on a single stream.
- Fix the measurement method in advance. Which instrument, which sampling point, which dilution, who reads the result. Agreement after the data arrives is not agreement.
- Blind the operator. Whoever records the numbers should not know which loop is treated. This is standard practice in clinical work and almost unknown in industrial pilots.
- Pre-register the success threshold. Write down what result would make you reject the technology before the trial begins. Ours is usually 15% on the target parameter with the control held flat.
Five questions for the next supplier you meet
- What is the specific energy in kWh/m³ at my duty point, and what head does that assume?
- Show the size distribution and the concentration, with the instrument named.
- What does your device not do? (If the answer is "nothing", end the meeting.)
- Will you accept an acceptance test written against my baseline rather than your catalogue?
- Which of your published claims would you retract if a customer tested it and disagreed?
Where this page comes from
ISO 20480-1:2017, ISO 20480-2:2018, ISO 20480-3:2021 · ISO/TC 281 committee catalogue (iso.org/committee/4856666.html) · Nanomaterials 12(13):2175, "On Some Aspects of Nanobubble-Containing Systems" — a review of stability models and the counterexamples against each · 2026 review of food-system nanobubble measurement (light-scattering limits) · our own ZetaView reports, six of which are linked from the product pages · GB/T 42842 series status per the national standards platform. Where we could not reach a primary source, the claim is marked as unverified rather than repeated.
Still want a proposal?
Good. Send the baseline data with the inquiry — three months of the target parameter, the water analysis, and the current energy bill. We will tell you what we think the realistic delta is before you spend anything.
Send baseline data → Run the payback numbers