- October 17, 2026
- Posted by: editor
- Category: RECD & Emission Control Devices
Imagine two emission reports for two different Retrofit Emission Control Devices. Both show a particulate matter (PM) reduction in the mid-70s, comfortably above India’s 70% requirement. Then a research group measures the same two devices for particle number (PN). One removes more than 98% of the particles by count. The other removes barely half.
Both mass results were correct. Both number results were correct. They simply answered different questions.
This article explains what particle mass and particle number each measure, why an RECD can score very differently on each, how the two tests are done, and when it’s worth asking for a particle number measurement on top of the standard mass test.
Mass and Number: Two Ways to Count the Same Smoke
Think of a bucket of mixed gravel and fine sand.
- Weigh it, and the gravel dominates. The fine sand barely moves the scale.
- Count the grains, and the fine sand dominates. There are vastly more sand grains than pebbles.
Diesel exhaust is similar:
- Particle mass (PM) weighs everything collected on a filter. It’s dominated by the larger particles, which carry most of the weight.
- Particle number (PN) counts individual particles in a volume of gas. It’s dominated by the smallest particles, which are far more numerous.
A device can remove most of the mass by catching larger particles while letting many tiny ones through, which gives high PM efficiency but modest PN efficiency. Or it can catch nearly everything, which gives high efficiency on both.
What Diesel Particles Actually Look Like
Diesel exhaust particles fall into recognisable size groups, usually called modes:
| Mode | Typical size | What they are | Share of number | Share of mass |
|---|---|---|---|---|
| Nucleation mode | Below about 30 nm | Mostly volatile droplets of condensed hydrocarbons and sulfuric acid, plus some tiny solid cores | Often the majority | Very small |
| Accumulation mode | Roughly 30–500 nm | Agglomerated soot (carbon) particles with adsorbed organics and sulfates | Large | Most of the mass |
| Coarse mode | Above about 1 µm | Re-entrained deposits from pipe walls, silencers and the device | Very few | Can be significant if present |
A nanometre (nm) is a millionth of a millimetre. Even the largest accumulation-mode particles are far smaller than a human hair, and almost all of them fall within PM2.5. See PM2.5 vs PM10.
Two points are important:
- The nucleation mode is mostly volatile. These particles form as hot exhaust cools and dilutes, so their number depends heavily on sampling conditions.
- The accumulation mode is mostly solid. These are the soot particles a filter is designed to trap.
How Particle Mass Is Measured
The standard method is gravimetric. A measured volume of exhaust is drawn through a pre-weighed filter, the filter is conditioned and weighed again, and the mass gain is divided by the sample volume. Results are reported as mg/Nm³ (concentration) or g/kWh (per unit of work).
This is the basis of particulate limits for genset engines in India and of the CPCB requirement that an RECD achieve at least a 70% PM reduction. We cover the method in smoke opacity vs particulate mass and RECD emissions data in mg/Nm³ vs g/kWh.
Strengths: directly regulated, well-established, captures the solid soot plus condensed organic and sulfate material.
Limits: dominated by larger particles and insensitive to very small ones. After a highly efficient filter, the mass collected can be so small that weighing uncertainty becomes significant.
How Particle Number Is Measured
Particle number is measured with particle counters, most commonly a condensation particle counter (CPC). The CPC grows each particle into a droplet large enough to detect optically, then counts the droplets.
For regulatory-style measurements, the sample usually passes through a volatile particle remover first. This heats and dilutes the sample to evaporate volatile droplets, so only solid particles are counted. Regulations also define a lower size cut-off, below which particles aren’t counted. 23 nm has been widely used, and newer approaches extend down to about 10 nm.
Other instruments, such as electrical low-pressure impactors and mobility sizers, measure the size distribution as well as the total count, which is useful for research and diagnosis.
Results are reported as particles per cubic centimetre (#/cm³), or per kilowatt-hour (#/kWh) when exhaust flow and power are known.
Particle number limits are already part of some engine regulations abroad. For example, the European Union’s Stage V rules for non-road engines include a PN limit for many compression-ignition engines between 19 and 560 kW, as summarised by DieselNet’s EU non-road emission standards page. In India, RECD compliance is currently assessed on particulate mass.
Strengths: very sensitive to the smallest particles; highlights differences between filter technologies that mass can hide; reacts quickly to small leaks.
Limits: results depend heavily on sampling and dilution conditions, and on whether volatile particles are removed. Different instruments and cut-off sizes aren’t directly comparable. There is no PN requirement for RECDs in India at present.
Why an RECD Can Score Differently on Each Metric
Full-flow wall-flow filters
In a wall-flow filter, every bit of exhaust is forced through porous walls. Particles across the size range are captured, and the very smallest are caught especially effectively by diffusion. Well-designed wall-flow filters typically achieve very high reductions on both mass and solid particle number.
Partial-flow and flow-through filters
These let part of the exhaust flow past the filtering surfaces instead of forcing it through. They can achieve good mass reductions, because they catch many of the larger particles, but solid particle number reductions are often lower and more variable.
Oxidation catalysts
An oxidation catalyst on its own doesn’t filter solid soot. It reduces the organic fraction of particulate mass by oxidising hydrocarbons. Its effect on particle number depends on whether those hydrocarbons would otherwise have formed nucleation-mode particles.
The role of fuel sulphur
Research comparing particulate control technologies has shown that fuel sulphur can change the picture considerably. One study of PN and PM reductions across several technologies found that with ultra-low sulphur fuel, full filter systems gave high reductions on both mass and number. Partial filters and oxidation catalysts showed higher mass-based than number-based reductions. With higher-sulphur fuel, the number-based results changed because of sulfuric acid nucleation particles. See the SAE paper on particle size distributions and particle number reductions for various PM control technologies. This is another reason to use consistent, low-sulphur fuel during testing. See how fuel and engine condition can confound RECD performance tests.
Worked Example: Three Technology Types, Two Metrics
The figures below are illustrative. They show how the same baseline exhaust can give different results by mass and by solid particle number, depending on technology.
| Technology type | PM before (mg/Nm³) | PM after | PM reduction | Solid PN before (#/cm³) | Solid PN after | PN reduction |
|---|---|---|---|---|---|---|
| Full wall-flow filter | 80 | 18 | 77.5% | 2.0 × 10⁸ | 3.0 × 10⁶ | 98.5% |
| Partial-flow filter | 80 | 22 | 72.5% | 2.0 × 10⁸ | 9.0 × 10⁷ | 55.0% |
| Oxidation catalyst only | 80 | 60 | 25.0% | 2.0 × 10⁸ | 1.6 × 10⁸ | 20.0% |
Two things stand out:
- The full wall-flow filter and the partial-flow filter look similar by mass, with only five points between them. By solid particle number, they’re far apart.
- The wall-flow filter’s PM reduction (77.5%) understates how thoroughly it cleans the exhaust. What remains after the filter by mass is partly condensable material that forms downstream, not solid particles slipping through.
Ways Particle Number Can Mislead
Volatile particles created during sampling
If volatile particles aren’t removed before counting, the number can depend more on the sampling setup than on the device. Removing most of the soot can even increase nucleation of volatile particles downstream, because vapours that would otherwise condense onto soot surfaces form new tiny droplets instead. A total (volatile plus solid) count taken after a filter can then look disappointing even when the filter works well. This is why solid particle number, with a volatile particle remover, is the more useful measure for judging a filter.
Different instruments and cut-offs
A count with a 23 nm cut-off and one with a 10 nm cut-off aren’t the same measurement. Neither are counts with and without volatile removal. Before comparing PN results, check that they used the same definition.
Dilution sensitivity
Particle counts are highly sensitive to dilution ratio and temperature. A small change in setup between upstream and downstream sampling can change the count substantially. See how dilution can mislead RECD before-and-after measurements.
Regeneration and spikes
During regeneration, and briefly after it, particle numbers downstream of a filter can rise considerably as the soot cake is removed and volatile material is released. A short PN measurement that happens to catch a regeneration won’t represent normal operation. See RECD testing during regeneration.
Ways Particle Mass Can Mislead
Blindness to the smallest particles
Two devices with the same mass reduction can leave very different numbers of ultrafine particles in the exhaust, as the worked example shows.
Measurement near the detection limit
After a very efficient filter, the mass on the filter paper can be tiny. Small weighing errors, blank contamination or background particulate become significant. See which instrument calibration records should accompany RECD tests.
Condensable material
Part of the mass downstream of a filter can be organic or sulfate material that condenses after the device, which a filter isn’t designed to capture. This can make the mass reduction look lower than the device’s actual solid-particle performance.
Why Particle Number Is So Good at Spotting Filter Damage
One of the most practical uses of particle number is detecting a small leak or crack in a filter-type RECD. A worked example shows why.
The figures are illustrative and use the full wall-flow filter from the table above. Suppose a small crack or failed seal lets 2% of the exhaust bypass the filter entirely, while the other 98% is filtered as before.
Particle mass after the damage:
PM after = (0.98 × 18) + (0.02 × 80) = 17.64 + 1.60 = 19.24 mg/Nm³ PM reduction = 1 − (19.24 ÷ 80) = 76.0% (down from 77.5%)
Solid particle number after the damage:
PN after = (0.98 × 3.0 × 10⁶) + (0.02 × 2.0 × 10⁸) = 2.94 × 10⁶ + 4.0 × 10⁶ = 6.94 × 10⁶ #/cm³ PN reduction = 1 − (6.94 × 10⁶ ÷ 2.0 × 10⁸) = 96.5% (down from 98.5%)
By mass, the change is only 1.5 points. That’s well within the normal uncertainty of a gravimetric test, so a mass test probably wouldn’t detect the damage. See how to interpret uncertainty in an RECD emission reduction claim.
By particle number, the downstream count has more than doubled, from 3.0 to 6.9 million per cm³. That’s a clear signal that something has changed.
The reason is that an intact wall-flow filter leaves so few solid particles downstream that even a small bypass of unfiltered exhaust dominates the count. In practice, a large rise in downstream particle number at a matched condition, together with a fall in differential pressure, is a strong sign of damage. See can RECD efficiency change between new and aged conditions.
Units and Reporting for Particle Number
PN results can be reported in several ways, and they aren’t interchangeable:
- #/cm³ at stated conditions: a concentration, comparable only if the dilution correction and reference conditions match
- #/kWh: a specific emission, which needs exhaust flow and power, like g/kWh for mass. This is the form used in some engine regulations.
- Reduction percentage: comparable across devices only if the same definition (solid only, cut-off size) and the same test conditions were used
Before comparing any two PN results, check:
- Solid particles only, or total including volatiles?
- Lower size cut-off (for example 23 nm or 10 nm)?
- Dilution ratio, and whether the result is corrected back to raw exhaust
- Reference temperature and pressure
- Load, warm-up and regeneration state during sampling
Which Metric Answers Which Question?
| Question | Better metric | Why |
|---|---|---|
| Does my RECD meet the CPCB requirement? | Particle mass | The requirement is stated as a PM (mass) reduction |
| How does my installation compare with its type-approval result? | Particle mass | Type approval is based on mass |
| Is the filter intact, with no cracks or bypass? | Particle number (or opacity for a rough check) | PN rises sharply with even small leaks |
| How do two filter technologies compare on ultrafine particles? | Solid particle number | Mass can hide large differences |
| What is the health relevance of the remaining exhaust? | Both, interpreted with care | Mass gives total burden, number gives ultrafine exposure |
| How many kilograms of PM did the device avoid? | Particle mass | Kilograms are a mass quantity; see estimating kg of PM avoided |
When Is It Worth Adding a PN Test?
A PN test adds cost and needs specialist equipment. It is worth considering when:
- The site is health-sensitive, such as a hospital, school or residential complex, and stakeholders ask about ultrafine particles. See RECDs for hospitals and RECD for schools.
- You are comparing technologies before a large fleet purchase, and mass results are similar.
- You suspect damage to a filter-type device. PN downstream of an intact wall-flow filter is very low, so a crack or seal leak shows up clearly.
- You want to make claims about ultrafine particle reduction in sustainability or ESG reporting, and need evidence beyond mass.
It is not a substitute for the gravimetric mass test where compliance is concerned.
How to Specify a PN Test
If you decide to add particle number, write these points into the test scope:
- Solid particles only, with a volatile particle remover operating at stated conditions
- Lower size cut-off stated (for example 23 nm or 10 nm)
- Same instrument and dilution set-up for upstream and downstream, or before and after
- Dilution ratio measured and reported for each sample
- Same load, stability and regeneration conditions as the mass test (see why generator load must be matched in RECD comparison tests)
- Results reported as #/cm³ at stated conditions and, where flow and power are measured, as #/kWh
- Calibration records for the particle counter and dilution system
- Size distribution data if available, which helps explain the result
Reading a Report That Includes Both
When a report gives both mass and number reductions, read them together:
- High mass and high number reduction: consistent with an intact full-flow filter.
- High mass but modest number reduction: consistent with a partial-flow design, or a possible bypass. Check device type and ΔP history.
- Modest mass but very high number reduction: check for condensable material downstream, and whether the mass result is near the detection limit.
- Number reduction lower than expected for a wall-flow filter: check whether volatile particles were removed, whether a regeneration overlapped the sample, and whether the filter is damaged.
Related Reading
- Can RECD efficiency change between new and aged conditions?
- How to interpret uncertainty in an RECD emission reduction claim
- Why RECD test results may differ from day-to-day site readings
- When should an RECD be retested after changes to the DG set?
- What is a diesel particulate filter (DPF) and how it helps
Frequently Asked Questions
Does India require particle number testing for RECDs?
No. The CPCB requirement for RECDs is based on particulate mass, at least a 70% reduction. Particle number testing is optional and is mainly used for technology comparison, diagnosis and health-related discussions.
Why would particle number reduction be higher than mass reduction?
A full-flow filter captures almost all solid particles, including the many tiny ones, so solid PN falls dramatically. Part of the mass remaining downstream is condensable material that forms after the device, which keeps the mass reduction lower than the solid PN reduction.
Can particle number go up after an RECD?
Total particle number, including volatile droplets, can sometimes rise downstream of a filter, because removing soot leaves vapours to form new tiny droplets instead of condensing onto existing particles. Solid particle number, measured with volatile removal, should fall sharply for an intact filter.
Is a handheld particle counter good enough for PN testing?
Not for testing an RECD. Raw exhaust is hot, wet and far too concentrated for handheld ambient counters, and plume measurements are diluted unpredictably. See can a handheld AQI monitor verify RECD performance.
Does a higher particle number reduction mean an RECD performs better overall?
Not on its own. A high solid PN reduction shows that the device captures small solid particles very effectively, which is a strength of full-flow filters. Overall performance also includes particulate mass (the compliance measure), CO and HC conversion, back-pressure, regeneration reliability and durability. Compare devices on all of these, under the same test conditions, and treat PN as one important part of the picture, not the whole of it.
Which is more relevant to health, mass or number?
Both carry information. Mass reflects the total particulate burden and is the basis of most air quality standards. Number highlights ultrafine particles, which are the subject of ongoing health research. A well-performing filter reduces both.
Choose the Test That Answers Your Question
For compliance, particulate mass remains the measure that counts. For understanding what’s left in the exhaust, and for checking that a filter is intact, particle number adds valuable detail. Aceget supplies CPCB type-approved Retrofit Emission Control Devices for DG sets from 25 kVA to 2500 kVA, and we can advise on which tests fit your site and stakeholders. See our certifications, read about how an RECD works, or contact our team.
[…] RECD particle number vs particle mass: what each test shows […]