Evidence · Emissions

Exhaust emissions — Japan and the Philippines

One measurement in our records was made by an accredited testing body: JATA in Japan, 2013. It shows CO down 30.4%, HC down 23.3% and NOx down 15.1% — and CO₂ essentially unchanged at −0.4%. Everything else we hold is a stationary snapshot taken at idle, which is a far weaker kind of evidence. This page publishes both, keeps them apart, and records the one NOx figure we withdrew because nothing here supported it.

Accredited third-party measurement Exhaust emissions

Fuel and emissions are not the same claim, and they do not come with the same evidence. This is what we can actually show on emissions — including the parts that do not flatter us.

−30.4%Carbon monoxide (CO), accredited measurement

−23.3%Hydrocarbons (HC), accredited measurement

−15.1%Nitrogen oxides (NOx) — the highest NOx figure in our records

−0.4%CO₂ — essentially unchanged in the same measurement

Exhaust emissions — at a glance
Testing body	JATA — Japan Automobile Transport Technology Association, an accredited testing body
Vehicle	Toyota E-MCV20W, 2,994 cm³, 1,550 kg, 1MZ engine
Date	31 October 2013
Application	30 cc applied to the intake air filter
Odometer	94,076 km before, 94,164 km after — 88 km between the two measurements
Measured	CO, HC, NOx and CO₂ in g/km, by JATA’s own procedure
Not measured	Durability. There is one before reading and one after reading, and no repeat months later
Everything else on this page	Stationary readings at idle, taken by workshops and by a regulator — indicative only

The result

Every measurement, not just the best one

Exhaust emissions — The accredited JATA measurement, 31 October 2013. Lower is better for all four.
Pollutant	Unit	Before	After	Change
Carbon monoxide (CO)	g/km	0.483	0.336	−30.4%
Hydrocarbons (HC)	g/km	0.09	0.069	−23.3%
Nitrogen oxides (NOx)	g/km	0.053	0.045	−15.1%
Carbon dioxide (CO₂)	g/km	253.8	252.7	−0.4%

Three regulated pollutants fell substantially. CO₂ did not — 0.4% on a single measurement is not a result, and it is shown in grey for that reason. This is the chart to look at before reading any CO₂ figure from us: where we quote one, it is calculated from a measured fuel saving, not measured at the tailpipe.

Stationary readings taken with the vehicle at idle, before and immediately after application. These are snapshots, not emissions results — read the section below the table before using any of these figures.
Vehicle	Where and when	Measure	Before → after	Change
Toyota Estima 2.4 petrol	Smile Cars, Shimonoseki · 2012	CO (ppm)	528 → 67	−87.3%
Toyota Estima 2.4 petrol	Smile Cars, Shimonoseki · 2012	NOx (ppm)	56 → 1	−98.2%
Toyota Estima 2.4 petrol	Smile Cars, Shimonoseki · 2012	CO₂ (vol.-%)	11.8 → 11.7	−0.8%
Daihatsu Mira petrol	Auto Shop Yamazaki, Mitaka · 2012	CO (vol.-%)	0.18 → 0.09	−50.0%
Daihatsu Mira petrol	Auto Shop Yamazaki, Mitaka · 2012	HC (ppm)	40 → 90	+125.0%
Daihatsu HIJET petrol	Auto Shop Yamazaki, Mitaka · 2012	CO (vol.-%)	0.43 → 0.09	−79.1%
Hino diesel #6006	SBS Zentsu · 2021	CO (ppm)	275 → 125	−54.5%
Hino diesel #6006	SBS Zentsu · 2021	NOx (ppm)	122 → 67	−45.1%
Hino diesel #6006	SBS Zentsu · 2021	CO₂ in exhaust (vol.-%)	3.6 → 2.8	−22.2%

The one regulatory outcome in our records: a vehicle that failed a statutory emissions test and passed it after treatment, at an accredited testing centre.
Vehicle	Test	Before	After	Result
Nissan Urvan 2.7 diesel	LTO statutory emissions test, Sucat Branch, Philippines (opacity limit 2.5)	FAILED	PASSED	Passed
Nissan Urvan 2.7 diesel	Re-check one month later	—	Still passing	Passed

Exhaust emissions — how it was measured

Method before result

JATA measured the vehicle on its own procedure as an accredited body, before the treatment and again after it. That independence is the whole point: it is the only exhaust measurement in our records that was not taken by us, by a distributor or by a workshop with an interest in the outcome.

The three regulated pollutants fell — CO by 30.4%, HC by 23.3%, NOx by 15.1%. CO₂ did not: 253.8 against 252.7 g/km is a 0.4% difference, which on a single measurement is indistinguishable from no change at all. We point that out ourselves because CO₂ is the number a fleet buyer cares about for reporting, and this measurement does not deliver it.

The same JATA report also records fuel efficiency at 9.3 against 9.4 km/L. We do not use that figure and it is not in the table above: 88 kilometres is far too short a distance for a meaningful consumption measurement. An earlier internal table of ours quoted −0.4% here, which was in fact the CO₂ change copied into the wrong row. It is corrected in our records and stated here rather than quietly dropped.

The stationary readings in the second table were taken with the vehicle idling, in most cases minutes after application. Idle readings respond strongly to engine temperature, to how long the engine has been running and to the state of the exhaust system, and a single pair of readings cannot separate those from a treatment effect. We hold them, so we publish them; we do not build claims on them.

The Philippine result is different in kind and stronger than any of the percentages: a Nissan Urvan that failed the statutory LTO emissions test passed it after treatment at the accredited testing centre in Sucat, and still passed on a re-check a month later. It is a pass or fail rather than a number, and a regulator applied the criterion, not us.

How we measure, and what makes a fuel trial credible →

Exhaust emissions — what this result does not prove

The limits of this measurement

We publish these next to the result rather than in a footnote. If a limitation would change how you read the number, you should read it at the same time as the number.

Known limitations

One accredited measurement is one measurement. A single vehicle, on one day, with one before reading and one after reading. It is the best emissions evidence we have and it is still a sample of one.
It says nothing about durability. Nothing in our records shows what the same vehicle measured three or six months later, which is exactly the period over which the fuel trials suggest the effect develops.
CO₂ barely moved in the accredited measurement (−0.4%). Any CO₂ reduction we describe elsewhere is derived arithmetically from a fuel-consumption change, not measured at the tailpipe — that is a calculation, and we label it as one.
A considerably larger NOx reduction than this was carried in our own ROI and ESG material until August 2026. We could not trace it to any measurement in our records, so we withdrew it rather than leave it standing. The highest documented NOx reduction we hold is the 15.1% on this page, and that is the figure we stand behind.
The stationary readings include a result that goes the wrong way: HC on the Daihatsu Mira rose from 40 to 90 ppm, an increase of 125%. An earlier internal note treated it as a probable measurement outlier and left it in a footnote. It belongs in the table, and it is a fair illustration of how noisy idle readings are.
Two NOx snapshots fall to 1 ppm and the CO on one Philippine vehicle to zero. Readings of essentially nil are not credible as effect sizes and we do not quote them as percentages anywhere — see the exclusions below.
For the SBS Zentsu readings we hold only a derived summary table, not the primary measurement record. They stay on this page as indicative figures, and we have asked for the raw records before they are used in any commercial material.
What we exclude entirely: the LTFRB demonstration readings of −100% for CO and NOx, taken in the street immediately after application. In the same dataset CO₂ on one of those vehicles rose by 5.7%. We publish neither as a result — the first because a reduction to zero is not a believable effect size, the second because it would be selective to report the rise while suppressing the method that produced it. Both are in our records and neither is evidence.
The ship study observed a visibly reduced exhaust plume but had no measuring instrument on board. It stays a qualitative observation and gets no number.

Exhaust emissions — what it means

Reading this evidence in context

For a fleet operator, the practical split is this: on local pollutants — CO, HC, NOx — we have one independent measurement showing real reductions, and it is the strongest emissions evidence we hold. On CO₂, we have almost nothing measured at the tailpipe, and any CO₂ figure you see from us is calculated from fuel.

That matters for reporting. If your obligation is a Scope 1 CO₂ number, the honest route is the fuel measurement and the standard emission factor, not a tailpipe claim from us. If your problem is a local air-quality or roadworthiness limit, the JATA measurement and the Philippine test result are the relevant evidence.

The most useful thing on this page may be the Philippine result, precisely because it is not a percentage. A vehicle that fails a statutory test and then passes it has crossed a threshold that someone else defined. That is harder to argue with than any figure we could publish ourselves.

What would move this forward is straightforward and we have not yet had it done: a repeat accredited measurement on more than one vehicle, with a follow-up after three to six months.

Exhaust emissions — FAQ

Questions operators ask about this measurement

Does NanoEFX reduce CO₂?

Not measurably at the tailpipe, on the evidence we hold. The one accredited measurement puts CO₂ at 253.8 g/km before and 252.7 g/km after — a 0.4% difference on a single reading, which is not a result. Where we quote CO₂ reductions, they are calculated from a measured fuel-consumption change using the standard emission factor for that fuel. That is legitimate arithmetic and we label it as arithmetic, not as an emissions measurement.

Have you ever published a larger NOx reduction than this?

Yes, and we withdrew it. A considerably higher NOx figure appeared in our own ROI and ESG material until August 2026. We could not trace it to any measurement in our records — not to this accredited one, and not to any of the stationary readings — so we removed it from the site rather than keep it while looking for a source. The highest documented NOx reduction we hold is the 15.1% measured by JATA.

Why publish the readings that show emissions going up?

Because they were measured. HC on one vehicle rose 125% in a stationary test, and CO₂ on one Philippine vehicle rose 5.7%. Both are in our records. Leaving them out while keeping the favourable readings from the same sessions would mean the table was selected rather than measured — and an evidence page that has been selected is worth nothing.

Are the stationary tests worth anything at all?

As indications, yes; as evidence, barely. An idle reading taken minutes after application is dominated by engine temperature and running time. The consistent direction across several vehicles is worth noting, and it is why we would like a proper repeat measurement. It is not worth a number on a brochure.

What is the strongest emissions evidence you have?

Two things, in this order. The JATA measurement, because an accredited body did it independently and the reductions in CO, HC and NOx are substantial. And the Philippine LTO test, because a vehicle that failed a statutory emissions limit passed it after treatment and still passed a month later — a regulator’s threshold rather than our percentage.

Exhaust emissions — where this data comes from

Provenance

Accredited measurement: JATA (Japan Automobile Transport Technology Association), 31 October 2013 — Toyota E-MCV20W
Statutory test: LTO Emission Testing Center, Sucat Branch, Philippines — accredited; re-checked after one month
Stationary readings: Smile Cars, Shimonoseki and Auto Shop Yamazaki, Mitaka (both 2012); SBS Zentsu (2021)
Recalculation: Every percentage on this page recomputed from the recorded before-and-after values
Held as: Derived summary tables for the JATA and SBS Zentsu figures; primary documents for the 2012 and Philippine tests
Open items: Primary JATA and SBS Zentsu records requested; repeat accredited measurement with a three-to-six-month follow-up not yet commissioned
Last checked: 15 August 2026

Prove it on your own fleet

The only baseline that matters is yours

Every trial on this site was run on someone else’s vehicles, duty cycle and baseline. A NanoEFX pilot measures yours — and we will help you design it with a control group so the result stands up.

Book a risk-free pilot All trial results → Calculate your ROI → Ask a question