October 31, 2026
Posted by: editor
Category: RECD & Emission Control Devices

Nobody looks at the ceramic honeycomb inside an RECD. It sits in a steel can, wrapped in a mounting mat, and works quietly for months. You start thinking about it only on the day it cracks. That is when you learn it is the most expensive single part of the device, that a crack lets soot slip straight through, and that the replacement has to match the model on your compliance papers.

So when one RECD quote says “SiC filter” and another says “cordierite”, the difference deserves more than a line in a comparison sheet. The silicon carbide vs cordierite choice decides how the filter copes with heat, how much soot it can safely hold, how it reacts to oil ash, how heavy the unit is, and what you pay now and later.

This article reads like a materials datasheet. For each property we explain what it is, how the two ceramics typically compare, and what that means on a DG set. Then we consolidate everything into tables, connect it to duty cycle, list the evidence to ask for and run an illustrative cost calculation.

One caution first. Each ceramic is made in many grades, with different porosity, wall thickness and cell density. The comparisons here are typical directions, not guarantees for any product. The datasheet of the exact filter offered to you is what counts.

The materials in brief

Both ceramics are used to make wall-flow filters: honeycombs with alternate channels plugged at opposite ends, so exhaust must pass through a porous wall and leave its soot behind. If you are unsure which filter type your quote describes, read our comparison of wall-flow and partial-flow filtration in RECD systems first, because the substrate question matters most for wall-flow designs.

Cordierite is a magnesium aluminium silicate. It began as a flow-through catalyst support and was adapted for filters. It is widely used in heavy-duty diesel applications.
Silicon carbide (SiC) is a hard, non-oxide ceramic. It became common as a filter material in light-duty diesel vehicles and is also used in heavier and stationary applications.
Aluminium titanate is a third option, used mainly in light-duty vehicle filters. You are unlikely to see it in an Indian RECD quote, so it gets only a side note below.

DieselNet’s page on wall-flow monoliths is good background reading.

Property 1: Maximum temperature

What it is. The temperature at which a ceramic melts or decomposes. In a filter, the practical ceiling is lower, because ash can react with the ceramic and form compounds that melt earlier.

How they compare. Cordierite melts at roughly 1,450°C, and DieselNet puts its practical maximum operating temperature at about 1,200°C, held down largely by ash reactions. SiC does not melt under any condition a filter will see; it decomposes far above 2,000°C.

On a DG set. Normal exhaust temperatures are well below both limits. The danger is an uncontrolled regeneration: a heavily loaded filter whose soot ignites all at once, pushing local wall temperatures far above exhaust gas temperature. A cordierite filter can develop a melted zone, usually towards the outlet end. An SiC filter is far less likely to melt, although it has its own failure mode, covered next.

Property 2: Thermal conductivity

What it is. How quickly heat moves through the material. High conductivity spreads a hot spot; low conductivity concentrates it.

How they compare. SiC conducts heat many times better than cordierite.

On a DG set. When soot starts burning in one region, an SiC filter spreads that heat into neighbouring walls, flattening the peak. In a cordierite filter, heat stays local and a hot spot can grow hotter. This is a main reason cordierite filters need tighter control of soot loading. Conductivity also helps an SiC filter warm and cool more evenly as the DG set starts and stops.

Property 3: Thermal expansion and filter substrate thermal shock

What it is. The coefficient of thermal expansion (CTE) is how much a material grows when heated. Thermal shock is cracking caused when different parts of one piece expand by different amounts at the same time.

How they compare. Cordierite wins clearly here. Its CTE is very low, a small fraction of SiC’s, so a large one-piece cordierite honeycomb can tolerate a big temperature difference across it. SiC expands considerably more and is also stiffer, which gives a large monolithic SiC block poor thermal shock resistance. The industry answer is segmentation, described under Property 8.

On a DG set. Thermal shock comes from rapid swings: a cold start straight into heavy load, a sudden load drop after a long full-load run, or regeneration starting on one side of the filter. Cordierite resists the stress through low expansion but must avoid extreme hot spots. SiC limits temperature differences through conductivity and relieves the remaining stress at cement joints. Neither is immune, so sites with frequent starts and stops should weigh this property heavily.

Property 4: Density and heat capacity

What it is. Volumetric heat capacity is how much heat a given volume absorbs per degree of temperature rise. A higher value makes the filter a better thermal sponge.

How they compare. SiC is denser and has higher volumetric heat capacity than cordierite.

On a DG set. This sets the soot mass limit: the maximum soot load allowed before regeneration must occur, chosen so that if it all burned at once, peak temperature would stay safe. A filter that absorbs more heat can safely carry more soot, so SiC filters generally tolerate a higher soot load per litre than comparable cordierite filters.

That margin matters most where regeneration is unreliable: standby sets running short weekly tests, or sets idling at low load. Our article on choosing an RECD for a generator running at low load explains why such duty loads filters quickly. The trade-off is that a higher heat capacity filter warms up a little more slowly, which can delay catalyst light-off on a catalysed design.

Property 5: Porosity and pore size

What it is. Porosity is the share of the wall that is open space; pore size is the typical opening diameter. Together they govern gas flow and particle capture.

How they compare. This depends more on grade than on material. DieselNet describes wall-flow filter materials as typically having porosity of 45 to 50 percent or more and mean pore size of about 10 to 20 micrometres. Both ceramics are made across such ranges, including high-porosity grades for catalyst coating.

On a DG set. Finer pores and thicker walls generally capture more but raise pressure drop. Higher porosity lowers pressure drop and leaves room for coating, but weakens the wall. Once a thin soot cake forms on the inlet walls, it becomes the main filtering layer, which is why good wall-flow filters of either material capture fine solid particles very effectively. Our sibling post asks whether RECDs can capture ultrafine particles from DG exhaust.

Ask for clean and loaded pressure-drop figures at your engine’s exhaust flow and compare them with the engine maker’s limit. Our explainer on how exhaust back-pressure affects a DG set covers why.

Property 6: Mechanical strength

What it is. How much load the ceramic carries before cracking, whether from bending, mat compression, vibration or impact.

How they compare. SiC is generally stronger and harder. Cordierite is adequate when well canned, but more vulnerable to rough handling, point loads and vibration once the mounting mat degrades. In both, strength falls as porosity rises.

On a DG set. Exhaust lines vibrate, and RECDs are often craned onto roofs or lowered into basements. A core cracked in transit may look fine from outside, which is why a baseline pressure-drop reading at commissioning matters. For roof jobs, see our guide to rooftop RECD installation, structural support and lifting, and never stand under a suspended load.

Property 7: Chemical durability and ash interaction

What it is. Exhaust carries incombustible ash, mostly from lubricating oil additives, plus traces of wear metals. At high temperature, some of it can react with the ceramic.

How they compare. Cordierite is the more sensitive material. Ash can react with it at high temperature to form glassy, lower-melting phases that lead to failure, which is why its practical ceiling sits well below its melting point. SiC is generally more robust against ash, though not inert: in oxidising exhaust it slowly forms a thin, normally protective silica surface layer.

On a DG set. Ash builds up in every wall-flow filter. DieselNet’s page on ash in particulate filters notes that ash adds flow restriction and that pressure drop alone does not reliably indicate how much ash a filter holds. An engine burning oil makes ash faster, and with cordierite, the risk of ash-ceramic reactions during a hot regeneration grows. Check engine oil consumption and RECD suitability before choosing, and treat persistent blue smoke as a warning sign.

Property 8: Manufacturing form, monolith or segmented SiC filter

What it is. Whether the filter is one extruded piece or many smaller blocks joined together.

How they compare. Cordierite filters are normally extruded and fired as a single monolith, which low CTE permits even at large diameters. SiC filters are usually segmented: small square honeycomb segments, often around 25 mm across, cemented together with an inorganic cement that typically contains ceramic fibres, then machined to shape and skin-coated. The joints act as stress-relief lines.

On a DG set. Cement joints take up some frontal area, slightly reducing open filtration area for a given diameter. Joint quality also matters, because a weak bond can become a leak path or crack starter. On a supplier visit, add “how do you inspect segment joints and skin coat?” to our list of questions to ask during an RECD factory visit.

Property 9: Weight

SiC is denser, so an SiC filter weighs noticeably more than a cordierite filter of the same size. On a stationary DG set this is usually minor, but rooftop installations and long exhaust runs hung from slabs need supports checked for the complete RECD weight. Ask for the weight of the whole device, not just the core.

Property 10: Cost

How they compare. Cordierite is generally cheaper to make: inexpensive raw materials and one-piece extrusion. SiC needs much higher firing temperatures, often in controlled atmospheres, wears extrusion dies faster, and adds segmenting, cementing and machining steps. For similar size and grade, SiC usually costs more.

On a DG set. The substrate is only part of an RECD price, alongside housing, coating, controls, installation and type approval; our overview of RECD cost in India explains the build-up. But the substrate dominates replacement cost, which is why failure risk matters.

Consolidated table: silicon carbide vs cordierite

Property	Silicon carbide (SiC)	Cordierite	Practical meaning for DG sets
Melting / practical ceiling	Does not melt in filter conditions; decomposes far above 2,000°C	Melts around 1,450°C; practical ceiling about 1,200°C due to ash reactions	SiC tolerates runaway regeneration far better
Thermal conductivity	High	Low	SiC spreads hot spots; cordierite concentrates them
Thermal expansion (CTE)	Relatively high	Very low	Cordierite resists thermal shock as a monolith; SiC needs segmenting
Volumetric heat capacity	Higher	Lower	SiC usually allows a higher soot mass limit
Porosity and pore size	Grade-dependent	Grade-dependent	Compare the actual grade’s datasheet
Mechanical strength	Generally higher	Adequate when well mounted	Handling and mat condition matter more for cordierite
Ash resistance when hot	Generally robust	More prone to reactions	Oil-burning engines raise cordierite risk
Form	Segmented and cemented	Usually one-piece	SiC joints must be well made
Weight	Heavier	Lighter	Check supports on roofs and long runs
Typical cost	Higher	Lower	Failure risk can reverse the first-cost advantage

Failure modes: which material is more prone and why

These are tendencies derived from the properties above. Good design and operating discipline can give either material a long life.

Failure mode	More prone	Why	Typical DG trigger	What you notice
Local melting	Cordierite	Lower melting point, low conductivity, low heat capacity	Large soot load ignites after long low-load running	Pressure drop falls sharply; smoke returns; glassy zone near outlet
Cracking from thermal gradients	Both; segmentation counters SiC’s high CTE	Stiffness and CTE in SiC; extreme hot spots in cordierite	Uncontrolled regeneration, cold start into high load	Lower than expected pressure drop; soot streaks on outlet face
Ash-ceramic reaction	Cordierite	Ash can form low-melting phases with cordierite	High oil consumption plus hot regenerations	Glazed, discoloured patches; plugged channels
Joint or cement failure	SiC	Only segmented filters have joints	Weak bonding, severe repeated cycling	Soot tracks along joint lines
Handling cracks or chipping	Cordierite somewhat more	Lower strength; relies on the mat	Rough handling, worn mat, vibration	Rattling or loose core, chipped face

Every one of these lets particulate bypass the walls, so it is a compliance problem, not just a maintenance one. Never run the DG set with the filter removed or bypassed to keep power on; that is a compliance risk that can lead to sealing and penalties. When a core does fail, our article on RECD filter core versus complete assembly replacement helps you choose the repair route.

How duty cycle should drive the choice

Neither ceramic is simply “better”. The right one fits how your DG set actually runs, so start with your run-hour and load records; our guide to generator load factor and duty cycle shows how to read them.

Where SiC’s tolerance earns its price

Heavy soot loading with unreliable regeneration: short test runs, long low-load idling, or older engines with poor combustion.
Unmonitored regeneration: no differential pressure alarm, or nobody acting on it.
Oil-burning engines: more ash, more risk of ash-ceramic reaction (though the real fix is the engine).
Remote or unattended sites: where a failure could go unnoticed for weeks.

Where cordierite is viable and cost-effective

Steady, moderate-to-high load: hot exhaust keeps soot burning continuously, often helped by a catalyst.
Good monitoring and disciplined maintenance: back-pressure alarms that someone acts on.
Healthy engine with low oil consumption.
Budget-sensitive projects where the duty genuinely suits it.

Catalyst coating changes the picture

A catalysed filter promotes soot oxidation at lower temperatures, often with NO2 help, so less soot accumulates. That helps both materials but matters most for cordierite, because it reduces the chance of a large load igniting at once. It does not remove the need to manage regeneration. Our sibling post on catalysed vs uncatalysed RECD filters explains the operating differences, and DieselNet’s DPF regeneration page covers the chemistry.

A note on aluminium titanate DPF substrates

Aluminium titanate was developed to combine low thermal expansion with better heat tolerance than cordierite, without segmentation. It is used mainly in light-duty vehicles. If an RECD supplier offers it, ask for the same evidence as for any substrate, plus field references in stationary engines.

Questions to ask the vendor, and the evidence to request

A material name on a quote is not a specification. Get answers in writing; our guide to reviewing deviations in an RECD technical offer shows how to record them.

Exact substrate? Material, supplier or grade, dimensions, cell density, wall thickness, porosity and mean pore size. Ask for the datasheet.
Segmented or monolithic? For SiC, how are joints and skin coat inspected?
Soot mass limit and protection? How does the design prevent overloading, and what must operators do when an alarm appears?
Pressure drop? Clean and loaded values at your exhaust flow, against the engine maker’s limit.
Coating? If catalysed, what minimum exhaust temperature does it need?
Warranty on cracking or melting? Many warranties exclude damage from uncontrolled regeneration or engine faults; see RECD warranty terms, parts, labour and exclusions.
Is this the type-approved configuration? CPCB type approval applies to a specific RECD model, tested by recognised agencies such as ARAI and ICAT. The certificate lists the model, not necessarily the substrate. Ask in writing whether the offered substrate matches what was tested and whether a later substitution would affect approval; our guide on how to read an RECD test certificate helps, and the Central Pollution Control Board is where to confirm current procedures.
Field references? Similar engine size, duty and run hours, and how many cores have been replaced.
Replacement supply? Lead time, price and whether the same grade will be available in five years.

Illustrative worked example: five-year expected cost

This example is illustrative. All prices and probabilities are assumptions chosen to show the method, not market data or failure statistics for any product.

Take a 250 kVA DG set (Aceget offers RECDs for 250 kVA DG sets), with two offers identical except for the filter core.

Assumptions:

Cordierite core: ₹1,20,000
SiC core: ₹1,80,000
Labour, gaskets, mat and transport per replacement: ₹40,000
Business cost of a failure (rental set or lost operations): ₹1,50,000
Five-year horizon, at most one cracked substrate

Cost of one failure:

Cordierite: ₹1,20,000 + ₹40,000 + ₹1,50,000 = ₹3,10,000
SiC: ₹1,80,000 + ₹40,000 + ₹1,50,000 = ₹3,70,000

Expected cost = core price + (failure probability × cost of failure)

Site A, steady prime duty with good monitoring (assumed probability: cordierite 10%, SiC 5%):

Cordierite: ₹1,20,000 + (0.10 × ₹3,10,000) = ₹1,20,000 + ₹31,000 = ₹1,51,000
SiC: ₹1,80,000 + (0.05 × ₹3,70,000) = ₹1,80,000 + ₹18,500 = ₹1,98,500

Cordierite is cheaper by ₹47,500.

Site B, standby duty with short low-load tests and no alarm follow-up (assumed probability: cordierite 40%, SiC 15%):

Cordierite: ₹1,20,000 + (0.40 × ₹3,10,000) = ₹1,20,000 + ₹1,24,000 = ₹2,44,000
SiC: ₹1,80,000 + (0.15 × ₹3,70,000) = ₹1,80,000 + ₹55,500 = ₹2,35,500

SiC is cheaper by ₹8,500, despite costing ₹60,000 more up front.

Break-even. At Site B, cordierite matches SiC’s ₹2,35,500 when ₹1,20,000 + p × ₹3,10,000 = ₹2,35,500. So p × ₹3,10,000 = ₹1,15,500, and p = 1,15,500 ÷ 3,10,000 ≈ 0.373. If you judge cordierite’s five-year cracking chance on that duty to be above about 37%, SiC is the better buy. The business cost of failure drives this heavily: for a hospital or data centre that needs a compliant set during GRAP restrictions, it may be far larger than assumed here.

Once the device is in service, our method for tracking RECD service cost per operating hour helps you check these assumptions against reality.

Safety when inspecting a substrate

Lock out and tag out the DG set before opening the RECD, and let it cool fully, as ceramic cores hold heat long after shutdown. Use lifting gear for cores and housings, wear gloves, eye and respiratory protection around soot and ash, and dispose of residues through a route confirmed with your SPCB. Photograph both filter faces so any crack pattern can be reviewed later.

FAQ

Is silicon carbide better than cordierite for an RECD filter?

Not automatically. SiC offers a far higher temperature margin, better heat spreading and more tolerance for heavy soot and ash. Cordierite offers very low thermal expansion, less weight and lower cost, and suits steady, well-monitored duty.

Why are SiC filters segmented?

SiC expands more and is stiffer than cordierite, so a single large block would crack under temperature differences. Cementing small segments together lets the joints absorb that stress.

What is the maximum temperature for a cordierite DPF?

Cordierite melts at roughly 1,450°C, but DieselNet gives its practical operating ceiling as about 1,200°C, mainly because of ash reactions. Runaway regeneration of an overloaded filter is the usual way it is exceeded.

Does substrate material change capture efficiency?

Capture depends mainly on filter type, porosity, pore size, wall thickness and the soot cake, rather than the material name. Compare the tested performance of the actual model.

Can I switch from cordierite to SiC at the next replacement?

Possibly, but not informally. Housing, mounting, pressure drop and controls may need changes, and the type-approved configuration may be affected. Get the supplier’s written confirmation first.

Does a cracked filter make my DG set non-compliant?

A crack lets particulate bypass the walls, so the device may no longer deliver its tested capture. Treat it as urgent, call your supplier, and never bypass or remove the RECD to keep running.

Ask for the substrate evidence, then decide

Aceget supplies CPCB type-approved RECDs for DG sets from 25 kVA to 2500 kVA. Aceget’s product page describes its RECD as using a silicon carbide ceramic filter with a precious-metal coating, an oxidation catalyst stage and NO2-assisted regeneration. Whatever material you choose, ask every supplier, including us, for the substrate datasheet, the soot-load protection approach and confirmation that the offer matches the type-approved model. Explore the Aceget RECD range, review our certifications, and contact our engineers with your DG nameplate, load log and run-hour history.

Silicon Carbide vs Cordierite in RECD Filter Substrates