Contour Enthusiasts Group Archives Forums General General Five Gas Smog Theory

It was suggested from the 3.0 Duratec Forum that an explanation of the gasses read in a smog test may be of interest. I will try to make a reasonably complete but brief explanation.

Going into the engine is Air and Fuel. The Air is composed of Nitrogen and Oxygen with a very small quantity of trace elements. The trace elements will not be delt with here. The Fuel is composed of Hydrocarbons, a mixture of many different molecules of different combinations of Hydrogen and Carbon. Other trace elements, such as Sulpher, will not be dealt with here.

In the combustion chamber, fuel is burned, which is also called oxidation. That means that Oxygen combines with these other elements. If combustion were perfect (it never is) you would end up with H2O and CO2. What you do end up with is some Oxygen that never combines (O, or more properly O2), Hydrocarbon that never combines (HC or raw fuel), Hydrogen dioxide (H2O or water), Carbon monoxide (CO) that didn't fully oxidize, Carbon dioxide (CO2), and Nitrogen (N). Normally N is inert and just passes through uneffected. If the combustion temperature gets too hot, the normally inert Nitrogen gets "scortched" and generates Oxides of Nitrogen (NOX).

So coming out of the combustion chamber you woould have O, HC, CO, CO2, H2O, N, and NOX. Of these, HC, CO, and NOX are considered pollutiants and are controlled to stay under certain levels by law. Of these 7 gases, H2O and N normally are not measured. The remaining five gasses can be helpful in diagnosing engine performance. More on that in a moment.

NOX is not generated until the combustion temperature is above (I think) 2500 degrees F. Normally the combustion temperature only gets this high when the engine is under load, hence the need to read NOX with the car on a dyno. NOX is reduced by retarding the timing, keeping the fuel mixture near ideal or slightly lean, altered cam timing (especially the exhaust valve) and exhaust gas recirculation. Exhaust gas recirculation (EGR) cuts combustion temperature because the exhaust gas that is sent back into the combustion chamber is for all intent and purpose inert to additional combustion. NOX can also be reduced after the fact with a reducing catalyst. Reducing means that it removes Oxygen, in this case from NOX. Reducing catalists are most effecient if the fuel mixture is held close to ideal. Don't confuse this with an Oxidizing catalist, which we will talk about in a moment. Any or all of these things may be used an any one engine design to clean up NOX.

HC and CO can be reduced by getting more Oxygen into the combustion process. This is done by continuing the oxidation after the combustion chamber while the gasses are still hot. Air injection (smog pump) is the oldest method, injecting air into the exhaust stream, usually immediately after the exhaust valve. Oxidizing catalysts encourage HC to become H2O and CO2 and encourages CO to become CO2. Oxidizing catalysts work best if the mixture is never lean, but either ideal or rich.

The need to keep the mixture closer to ideal so that both reducing and oxidizing catalysts can be used together was one of the main factors in developing technology for a closed loop fuel system. That is where oxygen sensors came in.

OK, now we are down to where we can discuss how reading the gasses from the tailpipe can be used for diagnosis.

HC is considered the universal ineffeciency measure. The higer the HC reading, the less effecient the engine is running. HC will be high is there is an engine miss. If the fuel is not burned, it remains raw fuel. HC will also be high if the engine is too rich or too lean. When the fuel mixture is near ideal (perfect stoichiometry), the HC reading will be at it's lowest, or the "trough" of a graph.

CO is referred to as the universal richness indicator. The richer an engine runs, the higher the CO will be. On a graph, CO increases as the mixture becomes richer. CO is not a dependable indicator of an extremely lean mixture because it is hard to read accurately as it appoaches 0.

O or Oxygen is considered the universal leaness indicator. The normal Oxygen content of Air is a little under 20% (about 18.6%). The reading from the tailpipe is the Oxygen over after oxidation. The higher the reading, the leaner the fuel mixture. When running properly, most engines will be at about .5% (without air injection). On a graph, Oxygen climbs as the mixture gets leaner.

CO2 is considered the universal effeciency indicator. The higher the CO2 reading, the more effecient the fuel mixture. Low CO2 can be from being either too rich or too lean. It can also be from a misfire or something else that hurts effeciency. On a graph, CO2 will be near it's peak with ideal fuel mixture (perfect stoichiometry). Technically speaking, the ideal mixture will be just barely on the lean side of the peak, but not much. Readings of 13.5% on cars with oxidizing catalysts are not uncommon.

So there you have it. It would be easy to write a book on emissions theory, but this is about as brief as i can make it and still have it make sense.

Big Jim wrote:

?Hydrogen dioxide (H2O or water)?

  ?Hydrogen dioxide?, if it existed, would be HO₂, and would be something entirely different than water.  

OK, so my High School Chemistry is a little weak. That was many years ago. H2O is correct, and it is water. Large quantities of water are generated from a running engine. Most of it is water vapr and is not observable.

There may be other errors in this hastily written offering, but try to look at the overall concept.

You go Jim, that was enlightening.

I had a chem class years back that asked how "dihydrogen oxide" content in the atmosphere was measured. The correct answer was relative humidity or something like that. He wanted to know if we could recognize water in chemical naming convention. Stupid trick question. Lots of people got that one wrong.

Anyway, thanks for the primer Jim. The engineer in me is happy to know what those measurements mean now. If I could only talk my mech into running my car on his e-test dyno after hours.

Good stuff, thanks.

Big Jim, very informative, but I must correct one thing: Normal room air (and outside, too) is 20.9 % Oxygen...
Sorry, I'm in the 02 business.

You're correct. Oxygen sensor in normal air usually reads 20.8 or 20.9. I got messed up about the Nitrogen to Oxygen ratio.

On reading the thing again, I also found an error concerning fuel mixture and NOX generation. NOX can be lowered by a slightly righer mixture as the rich mixture cools the flame. If I remember correctly, a reducing cat works best with a slightly lean mixture though.

To reduce the NOx you would need to run leaner. NOx is highest when lambda=1. As lambda is increased the NOx curve has a negative slope and therefore decreases. CO is lowest when lambda=1. As the air/fuel ratio becomes leaner then our CO levels go up slightly.

Some good points you mentioned Jim of how to reduce NOx. Some other ways being used in industrial engines is water injection into the cylinder during combustion and miller timing or closing the inlet valve early to name a couple

Holy old post the day before shut down

Anyway, nope -

leaner equals higher combustion temp, which means more NOx formation (burns nitrogen right out of the air).

Richer is cooler, but too rich means poor combustion & unburnt fuel which leaves more HC's.