Tuning Wiki

Workshop “carbon” covers more than one material. Soot from incomplete combustion, carbonised engine oil, and inorganic ash form in different places, and a process that removes one of them can leave the others in place. The temperatures and test engines cited below belong to those sources. They are not a single figure for every engine.

Three materials that get called carbon

Soot is combustible carbon left by incomplete combustion. It is the main solid caught by a diesel particulate filter, and it also travels in exhaust gas recirculation (EGR). Regeneration is oxidation of this soot. It leaves the inorganic residue behind.

Scanning electron micrograph of engine soot at 5,000 times magnification, with a 5 micrometre scale bar
Scanning electron micrograph of soot formed during operation of automobile engines, at 5,000×. The scale bar in the frame is 5 µm. This is the particle form of combustible soot, not a photograph of a port or a filter. Photo: Dr Anatoly, 14 May 2020, Wikimedia Commons. CC BY 4.0. Saved again as a JPEG for this page; dimensions unchanged.

Oil-derived deposits are what usually coat the back of an intake valve. Esaki, Ishiguro, Suzuki and Nakada concluded that intake-valve deposits come mainly from engine oil: light fractions evaporate from the hot valve, and the heavy fractions oxidise and carbonise into amorphous carbonaceous aggregates. The deposit tended to form where the valve surface was about 230 to 350 °C. Fuel sprayed onto the valve slowed that process, because the fuel dissolves precursors and cools the surface.

Stępień describes intake-valve deposits as amorphous porous carbon containing oxygen and nitrogen, plus calcium, zinc and phosphorus from the lubricating oil. Combustion-chamber deposits are a related carbonaceous material built from high-boiling fuel fractions, fuel additives and oil. In the work he reviews, detergent additives that reduce intake-valve deposits can increase combustion-chamber deposits, because the additive package itself is high-boiling.

Ash is the incombustible fraction. In a diesel particulate filter it is the metallic residue of lubricant additives, trace elements in the fuel, and wear and corrosion products. It stays after soot has been oxidised. The same split applies to a gasoline particulate filter: lubricant detergents are a main source of the ash, with zinc anti-wear additives, molybdenum friction modifiers, wear metals and fuel impurities also contributing.

Intake valves

On a port-injected petrol engine the injector sprays onto the back of the intake valve. That spray is a continuous wash. Kalghatgi ranks deposit levels as direct injection, then carburettor engines, then port injection, for that reason. Port-fuel detergents are aimed at deposits the fuel can touch.

On a gasoline direct-injection engine the fuel is injected into the cylinder. The back of the intake valve is no longer washed. Guinther and Smith, using a 2.0-litre turbocharged direct-injection car, found the deposit to be a mixture of engine oil, wear elements, unburned fuel and exhaust-gas contaminants. Lubricant formulation changed how fast it accumulated. The deposits can upset combustion, cause unstable running or stick a valve, and after mechanical removal they form again. Stępień traces the oil to the valve guides and to the positive crankcase ventilation (PCV) system; fuel reaches those valves mainly by blow-back when the valve is open.

Kalghatgi’s sources for the precursors are fuel and lubricant reacting with hot EGR, PCV vapour and blow-by. Surface temperature decides where on the valve the layer sits. Stępień’s survey of combustion-chamber surface temperatures puts cooler intake-valve surfaces near 120 °C, piston crowns around 150 to 230 °C, exhaust seats around 300 to 320 °C, and exhaust valves above 800 °C. Esaki’s 230 to 350 °C window is the range in which the oil layer carbonises, not a claim that every intake valve runs at one temperature.

Inlet ports and the inlet manifold

The same oil mist, blow-by and EGR gas that reach the valves also wet the port walls and the manifold. On a diesel the effect is larger, because cooled EGR is soot-laden exhaust mixed back into air that already carries crankcase vapour. The layer starts soft and oily and bakes into a hard coke that fills ports and seizes swirl flaps.

Soot nearly closing the passage of a diesel EGR and intake joint
Soot nearly closing a passage in the photographer’s diesel EGR system and intake manifold, after the EGR valve and manifold were removed. Photo: Moosealope, 9 October 2005, Wikimedia Commons. CC BY 2.0. Longest side reduced to 1,400 pixels for this page; no crop.

BMW service information B11 03 14 for the M57Y diesel in the E90 335d treats this as a repair. With the service lamp on and the engine running roughly, the fault pattern cited is air-mass sensor 3FF1 together with cylinder rough-running faults 4595, 459B, 4592, 4596, 4593 and 459C. The stated cause is excessive carbon in the intake. The repair cleans the intake valves in both the swirl ports and the tangential ports, and it cleans the manifold off the engine. After that job the injector zero-mass adaptations are cleared before the test drive, through the IRAP and programming process written in the bulletin, and the diesel control unit is programmed to the latest software. That adaptation step belongs to this BMW repair.

EGR valve, pipe and cooler

EGR deposits are not one coke. A review of cooler fouling identifies thermophoresis as the dominant way soot particles reach the cool wall: the particles are driven from the hot gas toward the colder surface. Hydrocarbons, water and acids condense in the same cooler and bind the layer. The layer raises pressure drop and cuts cooler effectiveness, so the gas returned to the intake is hotter and dirtier than the calibration assumes.

Cutaway of a clean EGR cooler tube bundle at a motor show
Cutaway of an EGR cooler at the 2006 Paris motor show. Exhaust passes through the tube bundle, which is the surface that fouls. This example is clean. Photo: Ton1-bot, Wikimedia Commons. CC BY-SA 3.0. Longest side reduced to 1,400 pixels for this page; no crop. Adaptations of this photograph must stay under the same licence.

A later chemical review, following Oguma’s classification, separates the shapes by gas temperature. Powdery dry soot is described above about 150 °C. Hard, slick deposits on the valve and pipes split into a thermosetting lacquer between about 80 and 150 °C and a thermoplastic lacquer between about 60 and 80 °C. Softer deposits sit nearer the intake. Hard layers are associated with oxygenated aromatic hydrocarbons polymerising on the cooled wall in the presence of acids. Cleaning the valve pintle does not clear a cooler whose internal tubes are already packed. A valve that sticks after the coke is gone has a mechanical fault as well as a deposit.

Combustion chamber and piston crown

Combustion-chamber deposits sit on surfaces the flame and the fuel actually touch: the piston crown, the chamber roof and the valve faces that open into the cylinder. Stępień summarises them as products of the highest-boiling fuel fractions, additives and oil. In the gasoline direct-injection work he reviews, lean stratified operation produced more of this deposit than rich operation, and direct-injection engines produced more than port-injected engines under the same conditions. He also lists the ways a running engine already removes deposit: oxidation and gasification, evaporation of the volatile fraction, and mechanical loss by abrasion, flaking and fuel wash.

These deposits can raise octane requirement and change emissions and heat loss. Fuel, and any induction cleaner that survives into the cylinder, can reach them. Blasting the intake port stops at the valve head and leaves the piston crown as it was.

Exhaust valves, manifold and turbine

Exhaust valves run hot enough that a thick oil layer of the intake-valve type is less likely to survive. Stępień’s temperature survey puts exhaust valves above 800 °C. Carbon still appears when the engine burns oil or runs rich, on the valve, in the port, through the turbine and in the downpipe. Intake-port blasting stops at the valve head. The exhaust port, the turbine wheel and the catalyst are outside that job.

Exhaust valve from a 1961 Volga M-21 with a dense grey deposit on the face
Exhaust valve from a Volga M-21 engine built in 1961, photographed at that engine’s first repair. The face carries a dense grey deposit. This is one old engine, not a direct-injection intake valve. Photo: DL24, 30 September 2017, Wikimedia Commons. CC0. Longest side reduced to 1,400 pixels for this page; no crop.

Particulate filters

A wall-flow filter catches soot and ash on purpose. Soot is removed on the vehicle by regeneration. A review of ash in particulate filters describes passive regeneration as catalytic soot oxidation at roughly 200 to 500 °C, and active regeneration as oxygen-driven oxidation near 550 °C. Ash remains either way.

DieselNet describes the ash as a wall layer or as plugs at the outlet ends of the channels. Either form shortens the useful channel, cuts soot-storage capacity, moves the remaining soot toward the front of the filter, and raises exhaust restriction. Extra restriction and more frequent regeneration both cost fuel. On a catalysed filter the ash layer also separates the soot from the catalyst, which makes passive oxidation harder.

Sintered-metal diesel particulate filter loaded with soot
Sintered-metal diesel particulate filter loaded with soot. The pleated element is a different construction from the square-channel wall-flow monoliths described above. Photo: Michael KR, 1 November 2015, Wikimedia Commons. CC BY-SA 4.0. Longest side reduced to 1,400 pixels for this page; no crop. Adaptations of this photograph must stay under the same licence.

A gasoline particulate filter follows the same soot-and-ash split. In a close-coupled filter study, calcium and magnesium detergents contributed more ash than the zinc and molybdenum additives. Wear metals and fuel impurities were in the ash as well. A nine-car fleet test found that a higher-ash oil put more ash in the filter and raised backpressure; the ash layer also acted as an extra filter membrane. Lower-ash oil reduced that backpressure rise. The filter is still required.

A three-way catalyst is a different component. Masking, melting or oil poisoning of a catalyst is a different failure. A carbon-cleaning session does not repair a catalyst.

Which method reaches which deposit

The method has to meet the deposit. A gas or a fine spray follows the airflow of a running engine. A blast nozzle cleans the surface it can see, with the intake valves of that cylinder shut and the shell extracted. A bath cleans a part that has been removed and immersed. Ash leaves a particulate filter when the incombustible residue is physically carried out.

MethodWhat it can reachWhat it is used forWhat remains
Fuel detergentInjectors, port-valve surfaces the fuel wets, combustion chamberKeeping port-injection intake valves and injectors cleanThe unwetted back of a direct-injection intake valve; filter ash
Liquid or foam induction cleanerIntake tract, then the chamber and the exhaust as the engine is runSoft, oily intake and chamber depositsHard baked valve and manifold coke; filter ash
Solvent on a removed partManifold, throttle, EGR body, swirl-flap housingCoke that can be scraped, soaked and rinsedValves still in the head; ash inside a filter
Ultrasonic bathPassages in a removed metal partEGR bodies and manifolds after the bulk coke is broken offElectronics and elastomers left in the tank; valves still in the head
Hot tank, hot water or steamRemoved manifold and similar metal partsThe mild-solvent hot tank BMW names for the M57Y manifoldA closed intake-valve face; the valve cover, which that bulletin keeps out of a pressure washer; filter ash
Walnut-shell blastingIntake port and the valve face the nozzle can seeHard direct-injection and diesel intake-valve carbonPiston crown, EGR cooler core, turbine, particulate filter
Dry-ice blastingSurfaces in the line of the nozzleAccessible surface deposits, without spent gritHidden coke and filter ash. A published match against walnut shell on glazed valves was not found
Hydrogen and oxygen inductionThe path through the running engine into the exhaustSold as a whole-engine decarbonising sessionOn the car tested below, measured power and fuel consumption were unchanged. Filter ash is incombustible
Filter regenerationSoot inside a particulate filterCombustible sootAsh
Off-car filter cleaningFilter channels, with the unit removedAsh, and soot a regeneration did not oxidiseA cracked, melted or oil-soaked substrate, which cleaning does not restore

Fuel detergents

Detergent in the petrol is the normal control for port-injection intake-valve deposits, and the European CEC M102E and M111 engine tests exist to measure that effect. The detergent travels with the fuel. On a direct-injection engine that fuel does not wash the intake-valve tulip, so a tank additive cannot do the job Esaki assigned to fuel spray. Stępień’s caution still applies on engines where the additive does reach the chamber: intake-valve control and combustion-chamber deposits can move in opposite directions.

Chemical induction, including foam

Induction cleaners are solvents introduced into the intake air, sometimes as a liquid spray and sometimes as a foam that is meant to cling to the port. TUNAP’s service sheet for product 133 is a concrete example of the spray form. It is written for petrol injection engines. The engine is warm, the cleaner is sprayed into the intake at about 2,000 rpm, the engine is stopped for 30 minutes to an hour, a second dose is sprayed, and the car is driven. TUNAP recommends an oil and filter change afterwards, which is a sign that the solvent can reach the sump. The product page describes it as a port-injection valve and combustion-chamber cleaner.

That class of product suits soft varnish and oily deposits on surfaces the mist can wet. The M57Y intake repair treats a blocked diesel manifold as a removed-part job: ultrasonic cleaning or a mild-solvent hot tank for the manifold and flaps, and walnut shell for the ports. Flooding a running intake with liquid can fill a cylinder. The dose belongs to the product sheet for that engine.

Foam is the same chemical approach in a form that sits on the surface for a dwell. An independent deposit-mass comparison that isolates foam from other induction solvents was not found for this page. Soluble deposits respond. Glazed carbon on a direct-injection valve is a mechanical cleaning job.

Solvent cleaning of removed parts

Once a manifold, throttle or EGR body is off the engine, the deposit can be scraped and dissolved. Bulletin B11 03 14 sends the intake manifold and swirl flaps, with seals, bolts and fasteners removed, for ultrasonic cleaning or a mild-solvent hot tank, and it requires every port and flap to come back clean. Intake seals are replaced. The throttle bore and the inner bore of the EGR valve are sprayed with a parts cleaner after excessive EGR buildup has been scraped. The spray stays off the electrical components. The swirl ports in the valve cover are wiped. The bulletin says not to put the valve cover in a pressurized parts washer, because the crankcase breather can be damaged. Solvent left in a passage becomes the next fault.

Ultrasonic cleaning

An ultrasonic tank cleans by cavitation in a heated liquid. It is a method for a part that has been removed. BANDELIN’s EGR application note uses a heated aqueous bath at 50 °C for 20 to 60 minutes, after coarse soot has been taken off with a plastic or wooden scraper, followed by a water rinse and drying. Their note is explicit that spray cleaners and fuel additives do not shift an EGR valve that is already encrusted.

Seals and the electrical actuator stay out of the liquid. BMW’s M57Y bulletin names ultrasonic cleaning as a suitable route for the removed manifold and swirl flaps, after the seals and fasteners are off. The valve cover is wiped, and is kept out of a pressurized washer. The bath does not reach intake valves that are still seated in the head.

Hot washing and steam

On a removed manifold, a hot tank is a real workshop process, not a guess. The M57Y bulletin names “a suitable, mild solvent parts cleaning hot tank” as the alternative to an ultrasonic bath for the manifold and swirl flaps. Hot water or steam on that removed part rinses deposit the solvent has loosened.

Introducing water or steam into a running engine is a different and older practice, aimed at the combustion chamber. Liquid in quantity will hydraulic-lock a cylinder. Steam in the intake does not wash the back of an intake valve that is shut. Hydrogen-induction machines are sometimes described as steam cleaning because burning hydrogen produces water. The machine and the measurements are in the hydrogen section below.

Walnut-shell blasting

Crushed walnut shell is the abrasive BMW specifies for intake-valve carbon. The shell is hard enough to break the deposit. It is blasted with the manifold off, through a port adapter, while a vacuum takes the spent shell away.

For the N54 and N54T, service information B12 02 12, October 2014, attributes drivability complaints and an illuminated service engine lamp to intake-valve and port carbon. If the misfire test plan returns “Evaluation of the fault pattern indicates coking in the intake ports,” the repair is removal of the intake plenum, repair instruction 11 61 050, and cleaning with the BMW Group Carbon Blaster, part 81 29 2 208 034. The bulletin points to an attached tool procedure for safety and operation; that attachment is not reproduced here. The media named in the bulletin is walnut shell graded 20/30 SAE, or 0.45 to 0.80 mm. Where the cleaning is the repair, intake gasket set 11 61 7 547 242 is fitted.

For the M57Y, bulletin B11 03 14, November 2014, uses the same blaster and the same shell grade. Each cylinder is cleaned at its top dead centre. The crank is locked with pin 11 6 080, the engine is not turned backwards, and the cylinders follow firing order 1-5-3-6-2-4. In that locked position the intake valves of the cylinder being cleaned are shut. Operating pressure is 6 to 8 bar. A vacuum adapter sits on the port and a shop vacuum runs during the blast. The trigger has an air-only stage and a stage that adds shell, and the bulletin says to alternate them so the port is not packed with media. On the swirl port the straight wand is used for about 5 to 10 seconds, then the side-outlet wand for about 10 to 15 seconds. Remaining coke in the tight bend of the port is loosened with a pick whose tips have been ground down so they do not gouge the aluminium, and the port is blasted again until the port and valve are clean. Shell left in the port is blown out with the air-only stage. The locking pin is removed before the engine is started.

The blast cleans the port and the valve face the nozzle can see. The piston crown, the EGR cooler core, the turbine and a particulate filter are outside it. Guinther and Smith note that the deposits form again, because the oil, ventilation and EGR paths are still there.

Dry-ice blasting

Dry-ice blasting drives solid carbon dioxide at the surface. Dzido and Krawczyk review the mechanism as three effects together: the cold cracks the layer, the pellet strikes it, and the pellet sublimates. The carbon dioxide leaves as gas, so the waste stream is the deposit rather than spent media. The same review records the limits of the process in industry: unsuitable settings can damage or polish the surface, pellets lose effectiveness as they age, and the work area has to be ventilated because carbon dioxide can build up in the air.

Small white dry-ice pellets sublimating on a dark surface
Dry-ice pellets, about 0.5 to 1.0 cm across, sublimating in air. Photo: Richard Wheeler (Zephyris), 2007, Wikimedia Commons. CC BY-SA 3.0. Saved again as a JPEG for this page; dimensions unchanged. Adaptations of this photograph must stay under the same licence.
Diagram of a dry-ice pellet striking a coating and lifting it from a substrate
Schematic of dry-ice blast cleaning. A pellet strikes a coating on a substrate, and the drawing shows the coating lifting off. It illustrates the blast. It is not a measurement from an engine. Diagram: Matt0322, 18 March 2014, Wikimedia Commons. CC BY-SA 4.0. Longest side reduced to 1,400 pixels for this page; no crop. Adaptations of this diagram must stay under the same licence.

That literature is industrial surface cleaning. A controlled comparison against walnut shell on glazed direct-injection valves was not found for this page. Dry ice is a line-of-sight method for deposits the nozzle can see. It does not empty ash from the channels of a particulate filter, and it does not replace the closed-valve vacuum extraction of the BMW shell procedure.

Hydrogen induction

A hydrogen carbon-cleaning machine electrolyses water and feeds the hydrogen–oxygen mixture into the intake while the engine runs. In the machine tested by Synák the electrolyte was a 30 percent potassium hydroxide solution in demineralised water. The gas is flammable, and the electrolyte is a strong alkali.

Synák’s 2026 measurements are the clearest published check of a one-hour commercial session found for this page. The car was a 2001 Škoda Fabia 1.4 with 230,000 km and EGR, a high-mileage port-injected petrol engine rather than a direct-injection valve-coke case. An HCC-5000 machine supplied about 1 kg of gas over an hour, split between idle and 3,000 rpm. On a chassis dynamometer, power and torque after the session sat inside the ±2 percent instrument band. Brake specific fuel consumption moved from 287.2 to 287.0 g/kWh, inside the fuel-flow meter’s uncertainty. Road fuel consumption was unchanged. Exhaust-gas shifts were small, and Synák notes they can also come from an hour of hot running through the catalyst. At the idle emissions test, hydrocarbons went from 114 ppm to 101 ppm, still over the 100 ppm limit that failed the car. Fitting an extra catalyst is what brought the reading down. Cranking current, used as a rough comparison of cylinder sealing, did not change. The study did not borescope the intake valves or weigh a deposit, so it does not measure valve-carbon removal. It does measure the power, economy and emissions claims attached to this service, and those claims were not borne out on this car.

Synák also records an earlier diesel experiment, which this page has not re-read in the original, in which the carbon-covered area of an exhaust valve and seat fell from about 65 percent to about 15 percent during an hour of the same sort of treatment. That is a report of deposit change on a hot exhaust valve. It is not a result for the back of a direct-injection intake valve, and it is not a result for ash in a particulate filter. Ash does not oxidise during regeneration, so a gas passed through the engine leaves it in place.

Particulate-filter cleaning

Regeneration is the on-car method for soot. Active regeneration needs the exhaust hot enough for oxygen to burn the soot, and many cars never hold that condition on short trips. Ash is still there afterwards.

Off-car cleaning is how ash is removed. Fox, Yang and Hunsicker cleaned 34 heavy-duty filters from highway trucks, using pneumatic air, thermal calcination and wet washing, alone and in combination. They weighed what came out and they measured pressure drop. Pressure drop could improve while a large amount of ash was still in the filter, so a lower restriction reading is not proof that the filter is empty. In that heavy-duty sample the off-board clean was associated with service around 100,000 to 200,000 miles. A passenger-car interval is the one in that vehicle’s maintenance specification, read together with the soot and ash figures the control unit reports.

Cordierite diesel particulate filter after mechanical cleaning, with open channel faces and staining at the rim
Cordierite diesel particulate filter after mechanical cleaning. The channel faces are open, with staining left toward the rim. This is not the soot-loaded sintered-metal filter shown earlier. Photo: Michael KR, 1 November 2015, Wikimedia Commons. CC BY-SA 4.0. Longest side reduced to 1,400 pixels for this page; no crop. Adaptations of this photograph must stay under the same licence.

Workshop practice on heavy-duty filters, as described by Rush Truck Centers, uses three operations. Compressed air dislodges loose ash. A controlled bake, in their description typically 500 to 600 °C, oxidises soot and oil the vehicle could not burn, after which air removes the remaining ash. An aqueous wash is used for ash and contamination inside the wall, and the filter is dried before it goes back on the vehicle. A substrate that is refitted wet can crack when it next sees exhaust temperature. The approved process is the one specified for that filter. A cracked, melted or oil-saturated unit is replaced.

On-car liquids injected through a pressure or temperature port are a lighter intervention. They affect material the liquid reaches and a following regeneration can oxidise. Packed ash plugs stay in the channels.

Choosing the job

Inspect the place that matches the complaint before choosing a machine. A borescope through the manifold shows whether a direct-injection valve is glazed, lightly varnished, or clean. A diesel with a sticking swirl flap or an EGR valve that does not move needs the part off and the passages measured. A particulate-filter fault needs the soot load separated from the ash load, and the substrate checked for cracks and melting, before a regeneration or a cleaning is sold.

The cleaning lasts only as long as the supply of deposit stays slow. Oil past the valve stems or the piston rings, a crankcase-ventilation fault, an EGR cooler that is already restricted, and a driving pattern that never regenerates the filter all put the layer back. Guinther and Smith recorded that direct-injection intake-valve deposits re-form after mechanical cleaning. Low compression from ring or valve sealing, a failing turbocharger, and a damaged catalyst are separate faults. Removing carbon leaves those faults in place.

Sources

Reference notes from the documents below. Applicability follows those documents. This page is not a substitute for the procedure for a particular engine, and hardware testing by this wiki is not implied.

The photographs are files that can be shared. Each caption names the photographer, the licence and what the frame shows. Copies used here were reduced or recompressed as the captions say, and were not otherwise altered. A freely licensed photograph of glazed direct-injection intake valves, crushed walnut-shell media, foam induction, a hydrogen-cleaning machine, or an ultrasonic bath was not found, so those methods are described from the written sources only.