- October 31, 2026
- Posted by: editor
- Category: RECD & Emission Control Devices
Two RECD quotations land on your desk for the same DG set. One describes a ceramic honeycomb with “alternately plugged channels”. The other mentions a “metal substrate with open channels” and promises it “will never choke”. The prices differ, the back-pressure figures differ, and the efficiency claims are worded very differently. The real question underneath is wall-flow vs partial-flow filter: two fundamentally different ways of separating soot from exhaust gas.
This article takes each design apart, follows the exhaust through it, and then sets the trade-offs against each other in pairs, because almost every advantage one architecture has is paid for somewhere else. We finish with the compliance test that actually decides the matter in India, an illustrative worked example, and the questions to put to any supplier before you sign.
The two architectures in plain terms
Both designs sit in the exhaust line downstream of the engine and both aim to keep particulate matter (PM) out of the chimney. The difference is whether every molecule of exhaust is forced through filter media or whether some of it is allowed to travel past.
What a wall-flow filter is
A wall-flow filter is usually a ceramic monolith, most often silicon carbide or cordierite, made of thousands of small parallel channels. Neighbouring channels are plugged at opposite ends, like a chessboard. Exhaust enters a channel that is open at the inlet but blocked at the outlet, so it has nowhere to go except through the porous wall into a neighbouring channel that is open at the outlet.
That forced passage is the defining feature. Essentially all of the gas crosses filter media at least once, which is why this design is sometimes called a “closed” or full-flow filter. DieselNet’s technical summary puts typical wall-flow monolith efficiency at roughly 70–95% of total PM mass, with higher figures for the solid fractions such as elemental carbon and for solid particle number. If you want the substrate side of this choice, our companion piece on silicon carbide vs cordierite filter substrates covers it.
What a partial-flow (flow-through) filter is
A partial-flow filter, also called a flow-through filter (FTF) or particle oxidation catalyst (POC), keeps its channels open from end to end. Instead of plugs, it relies on features that divert part of the flow into filtering media. Common constructions include metal foil structures with deflection features that steer gas into layers of sintered metal fleece, as well as metal foams, wire mesh and similar porous media, almost always with a catalyst coating.
Because the channels stay open, some of the exhaust never meets the media at all. It simply flows straight through. That is the “partial” in partial-flow, and it is also why the design is described as “open”. The DieselNet overview of diesel particulate filters draws the same line: full-flow filters treat the total exhaust stream, while partial filters capture particles from only a fraction of it and have much lower overall efficiency as a result.
A note on labels in quotations
Supplier literature does not always use these terms consistently. Some partial-flow devices are marketed simply as a “DPF”, and some deep-bed media blur the line between categories. If a quotation does not say plainly which architecture is inside, ask. The basics of how an RECD works apply to both, but the performance envelope does not.
Anatomy: following the exhaust through each design
The clearest way to understand the trade-offs is to trace a parcel of exhaust gas from the turbocharger outlet to the stack.
Path through a wall-flow filter
- Inlet cone. Exhaust leaves the pipe and spreads across the face of the monolith. Good flow distribution here matters, because a jet hitting the centre loads that zone faster.
- Inlet channel. The gas enters an open channel. A few millimetres in, it is already committed: the far end of this channel is plugged.
- The wall. Pressure pushes the gas sideways through the porous ceramic. On a clean filter, particles lodge inside the pore network (depth filtration). Within hours of running, a thin soot layer builds on the inlet channel surfaces and becomes the main filter itself (cake filtration), which raises efficiency further.
- Capture mechanisms at work. Inside the wall and the cake, the smallest particles are caught mainly by diffusion, because their random Brownian motion carries them into pore surfaces. Larger particles are caught by interception and inertial impaction.
- Outlet channel. Clean gas collects in the neighbouring channel, which is open at the downstream end, and exits.
- What stays behind. Soot accumulates on the inlet side until it is burned off during regeneration. Incombustible ash from lubricating oil additives and wear metals cannot be burned and gradually packs towards the plugged end of the inlet channels.
Path through a partial-flow filter
- Inlet. Exhaust enters open channels that run the full length of the element. There are no plugs.
- Deflection. Along the channel, shaped features or flow resistances push part of the stream sideways into the fleece or porous layer. The rest of the stream carries on straight down the channel.
- Media contact. The diverted portion passes through or along the media. Particles are caught by diffusion and interception, and soot is held in the fleece where a catalyst and NO2 can oxidise it.
- Bypass. The undiverted portion, plus any gas that skirts the media because local resistance has risen, leaves with much of its PM still entrained.
- Re-entry and exit. Treated and untreated streams merge before the outlet. What reaches the stack is a blend.
- What happens when the media fills. As the fleece loads with soot, its resistance rises and more gas takes the open path. The device does not plug. Instead, its capture efficiency falls. DieselNet’s page on particle oxidation catalysts notes that an unregenerated flow-through device gradually lets PM pass through rather than blocking, and that captured agglomerates can be re-entrained later.
That final step in each path explains most of what follows. A wall-flow filter that is overloaded announces itself through rising back-pressure. A partial-flow filter that is overloaded announces nothing; it quietly filters less.
The seesaw: wall-flow vs partial-flow filter trade-offs
Think of each pair below as a seesaw. Push one end down and the other rises. No single architecture wins every pair, so the useful question is which end your site, duty cycle and compliance obligation can tolerate.
Filtration efficiency ↔ Back-pressure
Wall-flow: Forcing all gas through the wall delivers high and consistent capture, but that wall is also a resistance. Clean back-pressure is moderate, and it climbs steadily as soot loads, falling again after regeneration. Size the element too small and the engine sees excessive back-pressure for much of its run time.
Partial-flow: Open channels give low and relatively flat back-pressure. The price is efficiency: because some gas bypasses the media, capture is lower and moves with load, flow rate and loading state.
What it means for you: if the engine has little back-pressure margin, a wall-flow filter must be generously sized rather than swapped for a less effective design. Our guide to how exhaust back-pressure affects a DG set explains why the OEM limit is the number to work to.
Robustness to overload ↔ Guaranteed capture
Partial-flow: This is the design’s headline advantage. If regeneration stalls, for example during weeks of light-load running, the device will not block and will not stall the engine. Back-pressure stays tolerable.
Wall-flow: If regeneration fails for long enough, soot keeps accumulating and back-pressure keeps rising. Left unmanaged, that can trip the engine on high exhaust pressure, and a very heavily loaded filter that then regenerates suddenly can see a sharp temperature rise that risks cracking the substrate.
The other end of the seesaw: the same feature that stops a partial-flow device from blocking is the reason its capture is not guaranteed. “Never chokes” and “always captures” cannot both be true of one open channel. A wall-flow filter that is monitored and maintained gives you capture you can rely on; the blocking risk is a management task rather than a reason to accept lower filtration.
Ultrafine particle capture ↔ Simplicity of media
Wall-flow: Diffusion inside the wall and cake is very effective for the smallest solid particles, so wall-flow filters typically achieve high reductions in solid particle number as well as mass. Our sibling post on whether RECDs can capture ultrafine particles goes deeper.
Partial-flow: The media that the gas does contact can capture ultrafine particles well, but the bypass fraction carries ultrafines straight through. Overall particle number reduction is typically lower and more variable.
In both cases, volatile droplets can nucleate downstream of the device as the exhaust cools, so neither design is a full answer to particle number. India’s RECD rule is written on PM mass, not particle number, at the time of writing.
Ash storage ↔ Service interval
Wall-flow: Ash has nowhere to escape, so it accumulates in the inlet channels, taking up filtration area and adding to back-pressure over the long term. Periodic ash cleaning is part of owning one. The DieselNet summary on ash in particulate filters notes that ash mainly comes from lubricant additives and that pressure drop alone is not a reliable indicator of how much ash is present.
Partial-flow: Some ash passes through the open channels and leaves with the exhaust, so ash-driven back-pressure rise tends to be slower. That sounds attractive, but the ash leaving the stack is particulate the device did not stop.
For either design, high oil consumption shortens the interval. See how engine oil consumption affects RECD suitability.
Regeneration dependence ↔ Consequence of failure
Both designs depend on regeneration. Most partial-flow devices rely heavily on NO2 made by an upstream or integrated catalyst to oxidise soot continuously at moderate temperatures. Wall-flow filters can be passive (catalyst and NO2), active (heaters, fuel dosing or burners), or a mix.
The difference is the consequence when regeneration underperforms. In a partial-flow device, efficiency drifts down silently. In a wall-flow filter, back-pressure rises visibly and a monitoring system can raise an alarm. Our sibling article on catalysed vs uncatalysed RECD filters covers what the catalyst changes, including the higher NO2 share at the outlet that catalysed designs of both types can produce.
Size and weight ↔ Margin
For the same exhaust flow, a wall-flow element generally needs enough filtration area to keep back-pressure within limits across its loading range, which drives volume. Partial-flow elements can sometimes be packaged more compactly because their resistance is lower. On a crowded rooftop or in a basement plant room, that can matter. Do not let packaging alone pick the architecture; check the general arrangement drawing for both options before deciding space is the constraint.
Purchase cost ↔ Lifetime cost
Prices vary by supplier, rating and scope, so treat any comparison as project-specific. A partial-flow unit may quote lower or higher depending on its metal media and catalyst loading. A wall-flow unit adds the cost of ash cleaning and, where fitted, ΔP monitoring. Neither quotation is complete until you add cleaning, consumables and the cost of a failed compliance test. The RECD cost guide for India lists the items to bring into the comparison.
Wall-flow vs partial-flow filter: comparison table
| Factor | Wall-flow filter | Partial-flow / flow-through filter |
|---|---|---|
| Gas path | All gas forced through porous walls | Open channels; part of the gas diverted into media, rest bypasses |
| Typical PM mass capture | High; DieselNet cites roughly 70–95% total PM for monoliths, higher for solid soot | Lower and more variable; depends on design, load and loading state |
| Solid ultrafine particles | Typically very high capture | Partial; bypass fraction escapes |
| Clean back-pressure | Moderate | Low |
| Back-pressure as it loads | Rises with soot, drops after regeneration | Relatively flat |
| If regeneration fails | Back-pressure climbs; can block if unmanaged | Does not block; efficiency falls silently |
| Ash | Retained; needs periodic cleaning | Partly passes through |
| Monitoring value | ΔP gives a clear loading signal | ΔP says little about efficiency |
| Typical construction | Ceramic honeycomb (SiC, cordierite) | Metal foil with fleece, metal foam, wire mesh |
| Meets ≥70% PM rule? | Depends on type-approval test; most designs aim well above | Depends on type-approval test; harder margin |
The compliance angle: what the 70% rule really tests
In India, the question of wall-flow vs partial-flow filter performance is not settled by brochures. CPCB’s procedure for emission compliance testing of retrofit emission control devices covers in-use DG engines up to 800 kW gross mechanical power, and it requires a type-approved RECD to achieve at least 70% PM capture efficiency. Type-approval testing is carried out by recognised agencies such as ARAI and ICAT. You can follow updates on the CPCB website.
What that means for the architecture choice:
- The rule is technology-neutral. It does not ban partial-flow designs or mandate wall-flow. Any device that passes its type-approval test within its approved scope is eligible.
- The margin differs. Wall-flow filters typically operate well above 70%, so normal variation in engine condition, fuel and load leaves headroom. Partial-flow designs typically sit lower and swing with operating conditions, so the gap between the tested figure and the threshold can be thin.
- The certificate is device-specific. A certificate applies to the tested model and the engine range, power band and conditions stated on it. It does not transfer to “similar” designs. Learn how to read an RECD test certificate and check whether the engine family and kW band cover your set.
- Verify it independently. Our guide to verifying that an RECD is CPCB tested walks through matching the certificate to the device in front of you.
There is no particle number or black carbon requirement for RECDs at the time of writing; check CPCB for any change. If your organisation reports black carbon for ESG purposes, the sibling post on what a particulate reduction claim includes for black carbon is worth reading, because architecture affects that figure too.
Duty-cycle matching: when the no-blocking feature looks attractive
Compliance comes first. Once you have two or more devices that are each type approved for your engine, duty cycle becomes the tie-breaker.
Where partial-flow’s character appeals
- Very low-load standby sets that run weekly test cycles at 10–20% load and rarely reach temperatures where passive regeneration is effective.
- Intermittent short runs, such as sets that start for a few minutes during grid switchovers.
- Unattended sites where nobody will respond quickly to a high back-pressure alarm.
In these cases, a device that cannot block is tempting. But the same low exhaust temperatures that stall wall-flow regeneration also starve partial-flow catalysts of the NO2 chemistry they rely on, so the partial-flow device may spend much of its life saturated, and therefore filtering at the low end of its range.
Where wall-flow is the stronger fit
- Prime and continuous duty, or standby sets that run long blocks at 40% load or more during outages.
- Sites under close regulatory scrutiny, such as Delhi NCR during GRAP stages, where measured performance needs to hold up.
- Owners who want a clear health signal. A ΔP trend on a wall-flow filter tells you when soot or ash needs attention.
Fix the duty cycle rather than downgrade the filter
Often the better answer for a lightly loaded set is load management rather than a weaker filter: a load bank during test runs, consolidating loads onto fewer sets, or a design with active regeneration support. Use DG load logs to assess RECD suitability and read the guidance on choosing an RECD for a generator running at low load before concluding that the architecture must change.
Illustrative worked example: same engine, two architectures
The numbers below are illustrative. They do not describe any specific product or test result, and the actual CPCB procedure defines its own test points, sampling and calculation method.
The setup
- One 500 kVA DG set whose engine falls within the CPCB procedure’s power range, with a measured baseline (engine-out) PM concentration of 90 mg/Nm³, on a consistent dry, reference-corrected basis.
- Device A: wall-flow, assumed to capture 88% of PM at both high and mid load.
- Device B: partial-flow, assumed to capture 60% at high load and 72% at mid load. (Partial-flow efficiency can rise or fall with load depending on design; here the higher flow at high load pushes more gas down the open path.)
- Engine OEM’s maximum allowable exhaust back-pressure: assumed 10 kPa.
Outlet PM concentrations
Outlet = baseline × (1 − capture efficiency).
| Case | Calculation | Outlet PM |
|---|---|---|
| Wall-flow, high load | 90 × (1 − 0.88) = 90 × 0.12 | 10.8 mg/Nm³ |
| Wall-flow, mid load | 90 × (1 − 0.88) = 90 × 0.12 | 10.8 mg/Nm³ |
| Partial-flow, high load | 90 × (1 − 0.60) = 90 × 0.40 | 36.0 mg/Nm³ |
| Partial-flow, mid load | 90 × (1 − 0.72) = 90 × 0.28 | 25.2 mg/Nm³ |
(For simplicity, the same 90 mg/Nm³ baseline is used at both loads. In a real engine, baseline PM changes with load, which is one reason load must be matched in RECD comparison tests.)
Against the 70% threshold
- Wall-flow: 88% at both points. Passes with 18 percentage points of headroom.
- Partial-flow, mid load: 72%. Passes, but with only 2 points of headroom.
- Partial-flow, high load: 60%. Falls 10 points short.
- Partial-flow, simple average of the two points: (60 + 72) ÷ 2 = 132 ÷ 2 = 66%. Below 70%.
The simple average is only a way to show the effect of variability. It is not how the official calculation weights results. The lesson is that a device whose efficiency swings across loads can pass at one point and fail at another, which is why the type-approval scope and test conditions matter so much.
Back-pressure comparison (illustrative)
| Condition | Wall-flow | Partial-flow |
|---|---|---|
| Clean, rated load | 3.5 kPa | 1.5 kPa |
| Soot-loaded, before regeneration | 8.0 kPa | 2.5 kPa |
| Headroom to 10 kPa OEM limit when loaded | 10 − 8.0 = 2.0 kPa | 10 − 2.5 = 7.5 kPa |
| Ash-loaded after long service, before cleaning | 9.0 kPa | 3.0 kPa |
| Headroom when ash-loaded | 10 − 9.0 = 1.0 kPa | 10 − 3.0 = 7.0 kPa |
Read the two tables together and the seesaw is plain. The partial-flow device is far kinder to the engine on back-pressure, with 7.0 to 7.5 kPa of headroom against 1.0 to 2.0 kPa. The wall-flow device is far more effective and comfortably compliant, but it needs sensible sizing, regeneration and timely ash cleaning to stay inside the OEM limit. For a real 500 kVA DG set retrofit, the supplier should show this back-pressure calculation for the specific engine.
Questions to ask before you choose
Put these to every supplier, in writing, and compare the answers side by side.
- Is the filter element wall-flow, partial-flow, or a hybrid? What is the substrate material?
- Which type-approval certificate covers this exact model, and what engine power range and conditions does it state?
- What PM capture efficiency was measured during type approval, and at which test points?
- How does efficiency change at low load and at full load, and what happens to it as the element loads with soot?
- What is the clean and loaded back-pressure at my engine’s rated exhaust flow, and how does it compare with the OEM limit?
- How does the device regenerate, and what exhaust temperature or load does that need?
- If regeneration does not happen, what does the device do: block, alarm, or lose efficiency? How will I know?
- Is ΔP and temperature monitoring included? If so, what are the alarm thresholds?
- How is ash handled, and what is the expected cleaning interval at my oil consumption and run hours?
- Does the catalyst raise the NO2 fraction at the outlet, and has that been measured?
- What warranty applies to efficiency, not just to the hardware?
- Can you show a reference installation on a similar engine and duty cycle?
Our guide to writing a technical RFQ for an RECD project shows how to make these answers mandatory in the bid format. If you are also comparing non-filter technologies, see filter-based vs electrostatic RECDs.
Safety and maintenance notes for either design
Whichever architecture you install, servicing involves hot surfaces, heavy elements and fine dust. Isolate the DG set and follow lockout/tagout before opening any housing; the OSHA guidance on controlling hazardous energy is a useful reference for the principle. Allow the exhaust system to cool fully, use lifting equipment rated for the element weight, and wear respiratory protection when handling soot and ash. Never remove or bypass the device to keep the generator running; that is a compliance risk and can lead to sealing and penalties. Our RECD maintenance guide covers routine checks for both designs.
FAQ
What is the main difference between a wall-flow vs partial-flow filter?
A wall-flow filter forces all exhaust through porous walls by plugging alternate channels. A partial-flow filter keeps channels open and diverts only part of the gas into filter media, so some exhaust bypasses filtration.
Can a partial-flow DPF block?
It is designed not to block completely. When its media saturates with soot, more gas takes the open path, so back-pressure stays moderate but capture efficiency falls.
Is a wall-flow filter always compliant with the 70% rule?
No design is automatically compliant. Compliance depends on the specific device passing type-approval testing within the scope shown on its certificate. Wall-flow filters typically have more headroom above 70%.
Does a flow-through filter reduce back-pressure on the engine?
Generally yes. Open channels give lower and flatter back-pressure than a wall-flow element of similar size, which is why some engines with tight back-pressure limits are offered partial-flow options.
Which is better for a standby set that runs at low load?
Check type approval first. After that, consider whether load management or active regeneration support can make a wall-flow filter work reliably, rather than accepting lower capture.
Does either design reduce NOx or CO2?
No. Neither architecture meaningfully reduces total NOx, and neither reduces CO2. Catalysed versions can raise the NO2 share of NOx.
Talk to Aceget About the Right Filter Architecture
Aceget supplies CPCB type-approved RECDs for DG sets from 25 kVA to 2500 kVA. Aceget’s product page describes a silicon carbide ceramic filter with a precious-metal coating, an oxidation catalyst stage and NO2-assisted regeneration, with particulate reduction of up to around 96% depending on DG condition and fuel. Explore the retrofit emission control device range, review the type-approval certifications, or send your engine data and load profile through our contact page for a back-pressure and suitability review.
Related Reading
- Silicon Carbide vs Cordierite in RECD Filter Substrates
- Can RECDs Capture Ultrafine Particles from DG Exhaust?
- Filter-Based vs Electrostatic RECDs: What Should Buyers Compare?
- Catalysed vs Uncatalysed RECD Filters: What Changes in Operation?
- RECD vs DPF: What Is the Difference?
- How to Read an RECD Test Certificate
[…] so some visible emission under heavy load may be normal for that design. Our comparison of wall-flow vs partial-flow RECD filtration helps set […]