Catalytic converters have come into common use with the millions of automobiles on the road these days, greatly reducing their emissions of unburnt fuel and NOx. Because the converter is outside the engine, the considerable amounts of heat they produce is waste.

After learning catalytic converters are also being used on indoor propane heaters and even wood stoves, could they be installed on jet flame cans?

Platinum/palladium catalytic heaters require only sufficient fuel vapor and a certain amount of heat to complete combustion of hydrocarbons and other byproducts to CO2, H2O, and N2.

As seen with woodstoves and propane heaters, the amount of fuel usage is significantly less than flame alone (which makes much more shorter wavelength visible light).

For example, per BTU, a catalytic heater running on gasoline (fumes) burns around 30% less fuel (observed by the writer) than even a ceramic plate propane heater (with a constantly lit pilot flame).

With the jet, added heat from the converter should increase thrust per pound of fuel burned. I dare say catalytic afterburners might be awesome.

Has any of this been tried?

  • $\begingroup$ I can't see a mechanism where you could use the heat from the catalytic reaction to create extra thrust. $\endgroup$
    – GdD
    Dec 28, 2021 at 15:28
  • $\begingroup$ @GdD Catalytic heaters give more complete combustion. Reference "jet exhaust" on Google (CO, NOx, soot, particulate matter). A catalytic surface enables these flame byproducts to be converted more completely to CO2, H2O, and N2. In addition to cleaner air, these reactions are exothermic. Ask any mechanic how hot a CC can get. The converter would be inside the engine, adding heat along with the flame can. $\endgroup$ Dec 28, 2021 at 16:58
  • $\begingroup$ aviation.meta.stackexchange.com/a/3493/20174 $\endgroup$ Dec 28, 2021 at 23:32
  • $\begingroup$ Catalytic converters, once warmed up, work so well that the flame can be extinguished and the converter alone will consume all of the fuel. A jet engine burns around 1 liter per second. $\endgroup$ Dec 29, 2021 at 1:06

3 Answers 3


As an addendum to Anonymous Physicist's answer:

Car type catalytic converters consist of mesh or honeycomb structures seeking to get the majority of the molecules in the stream exposed to the catalyst. By definition this prevents optimal supersonic flow since there must be interaction between fixed parts of the cat structure and the entire flow volume, reducing the available thrust.

This paper doing CFD on a car exhaust seems to be focusing on flow rates under 10 meters a second, vs several hundred for a jet engine, and seems to indicate that at that speed the needed length was 2 meters to allow reactions to complete. It would seem likely operating at increased flow velocity would require a longer catalyst bed.

Taken together this means that adding a catalytic converter to a jet engine it would involve massively increasing the flow area at or just downstream of the combustion area by a factor of between 10-100 reduce the skin drag by lowering speed while in the converter and keep overall system length sane, it would then have to constrict again to get useful final exhaust velocity which has performance impact as well.

The overall high frontal area shape from this would be problematic for a aircraft engine but might do something for a stationary power plant, assuming there is actually much partial combustion going on. Reciprocating engines as used in cars have partial combustion since there are areas of the cylinder where either fuel/air mix is not correct or where the thermal mass of the structure inhibits combustion. A gas turbine burner section does not need to perform the other three elements of a combustion cycle (compression, expansion and exhaust) so can optimise both mixing and temperature to achieve high combustion efficiency.

As an alternative to a catalytic converter bed, just lengthening the combustion area would appear to have a similar result of maximizing combustion completion and it is notable that power turbines like the LM2500 do not appear to have longer combustion zones compared to their aviation parents suggesting that the performance costs of messing with the high velocity flow pushes increase in combustion completion into diminishing returns territory.

  • $\begingroup$ Better answer. What is the flow rate in the flame can? Is it anywhere close to supersonic? "Overall high frontal shape" was never a problem for centrifugal compressors, or, indeed fans. We need some hard data on conditions in the flame can, removing NOx alone would be a very positive step forward. I'm not buying "thrust loss" quite yet, as pressure drop begins at the turbine. $\endgroup$ Dec 29, 2021 at 5:20
  • $\begingroup$ @RobertDiGiovanni - some envelope math suggested a cat chamber 10 meters wide by 20 long, need some big gains for that to be worth while but also do not trust the numbers. I think the bigger problem is in the last para - I could not find any mention of unburned products in jet exhaust when I went looking at papers, the current focus seems to be NOx control by avoiding hot spots. $\endgroup$ Dec 29, 2021 at 5:34
  • $\begingroup$ That might definitely help commercial flights in our "good" ozone layer. $\endgroup$ Dec 29, 2021 at 5:38

Ignoring drag losses, the force generated by a jet engine is (mdot fuel + mdot oxidant)*exhaust velocity. (mdot is mass flow rate: mass per time)

If the catalyst retards the velocity of the exhaust jet, then you are just killing the force generated by the jet engine. And when I think of a catalytic converter, I think of long, narrow channels and lots of pressure drop.

Also, since the active surface of the catalyst is only doing the job of catalyzing the oxidation, the larger you make the engine, the worse the pressure drop will be per volume of gas. This is a classic surface area to volume scaling gotcha.

  • $\begingroup$ Pressure drop is at the turbine, not in the flame can. $\endgroup$ Dec 29, 2021 at 5:22
  • 1
    $\begingroup$ @RobertDiGiovanni Wait 'til you put that catalytic converter in. $\endgroup$
    – m_a_s
    Dec 29, 2021 at 15:34
  • $\begingroup$ agree that an auto style may significantly impede flow. The combustors on wood stoves look more like round waffles. I have seen them scrub a propane flame, and (once fully warmed) handle all of the propane flow (with the flame out). Consider that the catalytic rxn is faster than flame. So, a platinum/tungsten woven mesh at the end of the flame can would best Test # 1. $\endgroup$ Dec 29, 2021 at 18:35
  • $\begingroup$ How is the catalytic rxn faster than a "flame"? I would like to see some data. Flames only need to mix and react. While the catalyst may increase the rate of reaction while it is on the catalyst, overall it is slowed by the time for adsorbing onto the catalyst, desorbing from the catalyst, and that's after the reactants have to diffuse through the materials surrounding the catalyst in the first place. Since you want to use the "waffles" which have a relatively low surface area, you are not going to burn everything on the catalyst. $\endgroup$
    – m_a_s
    Dec 30, 2021 at 16:53
  • $\begingroup$ "flames only need to mix and react" at a much higher temperature. The dwell time for combustion is very short and the mixture (because of turbulence) is not consistent within the flame. Platinum is on the surface and very fast. See 3rd comment. $\endgroup$ Dec 30, 2021 at 17:22

A fluid catalyst would not make any sense. It would get ejected from the engine, defeating the point.

A solid catalyst could improve efficiency, but the improvement would be minuscule. Solid catalysts work at the surface. Compared to the mass flow through a jet engine, the surface available to be coated with catalyst inside the engine is extremely small. This ensures poor catalyst performance. Considering that catalysts add weight and require periodic replacement of precious metals, this is not an economically plausible idea for aviation.

You cannot increase the surface area of the jet engine or its exhaust nozzle because there would be a large drag penalty. Jet engine shapes are highly optimized.

This might make sense if you were using the jet engine as a snow melter on the ground. That's a rare situation. Usually it is better to move the snow instead of melting it.

  • $\begingroup$ Just to help a little, the cat(s) would be in the high pressure, lower flow section, on or near the flame cans. $\endgroup$ Dec 28, 2021 at 22:57
  • $\begingroup$ I think the answer made it quite clear in paragraph 3 why you cannot use a powder or honeycomb structure in a jet engine. $\endgroup$ Dec 28, 2021 at 23:16
  • $\begingroup$ If most of the mass flow is not involved in combustion because it's nitrogen or excess oxygen, then the catalyst won't work on it. $\endgroup$ Dec 28, 2021 at 23:17
  • $\begingroup$ "the surface available to be coated with catalyst inside the engine is extremely small." Yes well the surface area inside a typical automatic exhaust pipe is quite small too, but they still manage to make it work by using a honeycomb structure (e.g. s.hdnux.com/photos/77/25/50/16603680/3/rawImage.jpg). You could do the same in a jet engine, but there would be a big weight penalty, which is a bigger deal for plane than a car, and also there would be a pressure drop across the structure, which would hurt performance. I don't know whether it would be a net win or not. $\endgroup$
    – Daniel K
    Dec 29, 2021 at 1:48
  • 1
    $\begingroup$ @RobertDiGiovanni do you have any maths for the increased drag/lost thrust from sticking a high area cat matrix in the exhuast stream? it looks like for cars adding the catalytic converter is around 5% performance loss due flow restriction, but cars are not depending on exhaust flow rate for thrust. $\endgroup$ Dec 29, 2021 at 1:54

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .