Failed MOT due to high CO Emissions

[comm_SS_V8]

A little help/advise please.

The car has failed it's MOT on just the CO (carbon monoxide) output. Everything else was OK. The limit set by the MOT Tester was 0.3%, the actual emissions on 2 Fast Idle and 1 Normal Idle test runs were 0.44, 0.44, & 0.45 respectively.

The engine was warm during testing. It also had Tesco 99RON fuel in so it should have been burning optimally. I have new O2 sensors, exhaust system and 200 Cell Cats. IIRC the factory default cats are around 400 Cell so there is a potential for the cats to let more emissions through.

Just doing a little research in an attempt to understand why/how the CO is high...

According to Haynes:

haynes said:

The MOT tester has access to a DOT database or emissions handbook, which lists the CO and HC limits for each make and model of vehicle.

I doubt very much that he looked up the DOT values for this car so I assume he used some 'default' setting. 1st Q, is the 0.3% setting OK? Could someone confirm what values should be used when testing emissions for these cars/LS1 engines please.

I've just put some RedX fuel system cleaner in the tank and will be driving North (200miles) tomorrow so it should flush it through. I'm also going to get the car hooked up to EFILive tomorrow to veryify she's not overfueling and that the maps are correct. The car's booked in for a re-test on Sat morning so I hope to have it fixed by then. If anyone has any ideas as to cause & rectification I would be most grateful.

[jagsy]

Perhaps need some work on the tuning??????

[caspy]

Fast Idle Test CO = 0.3% max

Natural Idle Test CO = 0.5% max

( for cars registered before 31st Aug 2002)

Both i believe are tested.

So his testing criteria were spot on. have just had 100 cell cats installed and passed quite easily. Must be a mapping/fueling issue. Have a word with Paul at Wortec.

>> Edited by caspy on Thursday 15th December 14:46

[ringram]

Hi Steve,

According to my recent researching CO is minimised at around 15.2:1 or better AFR, though you should be good at stoich, which the O2 sensors should be placing you at. So thats weird.

Basically HC is minimised at stoich and NOx at 11.5:1 then again around 18.x:1 (too lean)
Anyway its strange your CO is high at idle as the O2 sensors should keep you at stoich (14.63:1)

However I know you have recently had long tubes installed and if the o2 sensors are sited further down the exhaust and/or get colder their readings get thrown off. Maybe making the engine dump more fuel in.

If this is whats happening there are a few things you can do;
- wrap your headers with heat shield wrap
- sort the tune out to change the o2 sensor lean to rich trigger point
- use different o2 sensors (some guys use corvette rear sensors, they have 2 sensors on each bank, one pre and one post cat)

Whatever happens let us know how you go.

[stevieturbo]

Sensor heat/location shouldnt be an issue. The sensors are heated, so if they are working, it should be good.. Unless somewhow the cats arent up to temperature, but if the engine is at operating temperature, then this should be fine too.

Is the ecu std ? or has it been remapped ?

[gregwatson]

High CO means mixture too rich I think. Although the O2 sensors (assuming it's warm and running closed loop) should, as has already been said, keep the AFR correct. Unless it's too far out or unless the O2 sensors are reading incorrecly lean, causing the PCM to richen the mixture? Or possibly leaking injector(s), fuel pressure too high?

[comm_SS_V8]

Many thanks for the info chaps, it's greatly appreciated and informative. I'll have the maps & fueling checked first thing in the morning.

Will let you know how I get on.

[ringram]

Yes high CO definately means too rich.
Sensor location might matter as the heaters only get them started, they need to be about 350 degrees C to read accurately, colder than that and they wont read right. This is common with replacement headers as the tubular manifolds disipate heat faster.
Also if you have an exhaust leak that can cause the sensors to read "lean" when its actually not and tell the pcm to dump fuel.. again high CO

All easy fixes, you just need to know which one it is.

[caspy]

.....or the cats could be f***ed!!!

[gregwatson]

If it was me, I'd use either a voltmeter or preferably an OBDII scanner to watch the O2 sensor voltages. That's how I traced an exhaust leak on my kit car because I could see that the AFR at the sensors was perfect, and yet by the time the gases reached the tailpipe they were failing MOT on lambda too high (ie too much O2). Only explanation was an exhaust leak.

[caspy]

[quote]If it was me, I'd use either a voltmeter or preferably an OBDII scanner to watch the O2 sensor voltages. That's how I traced an exhaust leak on my kit car because I could see that the AFR at the sensors was perfect, and yet by the time the gases reached the tailpipe they were failing MOT on lambda too high (ie too much O2). Only explanation was an exhaust leak.[/quote]

Car is failing on CO not lambda, unlikely to be a leak.

[gregwatson]

Yeah, I know, I was just using the leak as an example. But watching the O2 sensors will show if the AFR is at least correct leaving the engine...

[A57 HSV]

Going for an MOT at 3.00p.m today. I'll make some enquires!

[comm_SS_V8]

Dropped in to WRC this morning on my way North and was advised that everything appears to be running normally. The long fuel trim is +1% (is this ok??) and the O2 sensors were reporting correct voltages/readings. Unfortunately WRC don't have any emissions testing equipment onsite.

I have found a friendly MOT test centre who is willing to put the car through the emissions test first to see if it'll pass before I spend the £40 on the MOT test itself. It's good to be home I'll also take her on a run before the test to make sure she's absolutely up to temp.

If it fails again further investigations will be required but I have my fingers crossed that the long run & engine/injector cleaner may reduce the values enough to pass. A long shot I know. If not, then I guess another detailed look at the maps & cats will be called for. Thanks to Caspy it appears that I actually passed the NATURAL IDLE tests

[comm_SS_V8]

As expected the car failed again on the CO emissions.

However this time I was able to watch the screen as the emissions test was being carried out. Interestingly the CO values fluctuated quickly & wildly between +0.3% and +0.9% as the engine revs remained steady at approx 2700rpm. The HC outputs remained steady throughout at 15 (+/-1) and the Lambda readings were also constant at 1.005 (+/-0.002). Potential thoughts now is a bad sensor as I would not expect this crazy behaviour from static maps. Any thoughts?? TIA

[ringram]

Unless you get a wideband reading you wont know for sure. WRC should be able to tell you what the tailpipe AFR reading is. If its not leaner than 14.63:1 then Id say something is amis as tailpipe readings tend to be leaner than pre cat. O2 sensors are just swtiches and are temperature dependant so I wouldnt rely on them to be accurate unless all the prerequisites are met and even then only at stoich (no air leaks and hot enough to read correctly). Long term trims of 1% are fine.

If you wanted to you could tune open loop just in the idle range and make the engine lean at idle and ignore O2 sensors, that should get you though, but will require some messing about with a tuning product.