Hiya guys,

Just put my 55plate sti type uk in for its mot and everything was fine except emmmissions and failed on high co levels, got an additive for the petrol and ran it for a week and put it in for a retest and it failed again however the co did drop but not enough. The car does lie not getting used much as I car share to work (not sure if being idle could cause this), before the retest I did give it a hard drive. Not sure what to do next, I believe it still has the 2 oem cats fitted although not 100% (will ask garage if they looked), Could it be possible that cats have failed? the car has only done 40k miles,

Readings for the test is below:-

Fast Idle Test

CO level at 2485 = 1.06% Max is 0.20%
HC level at 2485 = 47ppm
lambda at 2485 = 0.99


Second Fast Idle Test

CO level at 2531 = .99% Max is 0.20%
HC level at 2531 = 27ppm
lambda at 2531 = 0.99

Natural Idle Test

CO level at 781rpm = 0.82% Max is 0.30%

I have a obd2 interface and there are no errors standing at all, however I am not sure if the readings displayed are within tolerance.

Any advice is most certainly appreciated
Cheers

 

excessive CO is caused by a rich a/f mix, i suggest you probley start some where checking things that can have a bearing on this. some sort of air leak maybe, in the intake system somewhere.

 

o2 sensor is a possibility hard to diagnose with out a decent diagnostic machine though

 

thnx for reply m8,
I did notice that the plastic ducting just as it enters turbo seemed to be a bit slack and not very tightly clamped, I would have tried to tighten it but its pretty awkward to get into it properly. Will deffo have another try at that tommorrow and let you know how i get on. Forgive me if i am wrong but would an air leak at the inlet especially if its after the airflow sensor mean excess air causing a leaner mixture hence reducing CO, again I could be completely wrong on this idea?
thnx again

 

No, the o2 sensor would pick up on the extra air and add more fuel

 

Do you have any mods? And yes you are right an air leak would show a lean mixture, my bad sorry! Could still possibly be a leaky vacuum pipe, just remind me on a 55 plate they are a closed loop lambda circuit yeah ? ( 2 sensors)?

 

I bought the car from a subaru garage and it was advertised as a ppp sti type-uk, only thing i can see on the car is a prodrive back box,oil filler cap and front mud flaps lol, so if any mods then prodrive is it. As for the lambda closed loop circuit sorry I aint sure about that. Tommorrow I will check all vacuum hoses and that intake see if if I can find anything.

Something a bit off subject While trying to find out info about my car I looked on the subaru main site then went to the prodrive option and downloaded the pdf file and i couldnt believe it, it is my car on the picture same reg ect lol, I phoned prodrive up about my beloved scooby but was told they never fitted pro drive the car it was used for photo shoots only, does that mean my car is famous lol

 

You are half correct, whilst EXCESSIVE CO is almost certainly going to be caused by a rich mixture, the figures shown above aren't that high and are not indicating that ( Lambda of 0.99, 1 being spot on).

It does look like a poor cat, (If as you have said they are in fact still there)also note the CO drop on the second test, you may find that running it for longer may cause it to eventually get hot enough to come good.

 

The HC reading looks like a sports cat is fitted, I've just MOT'd mine with 100cell in and HC was 52ppm.
I did take mine out for 20mins before the MOT to make sure it was well warmed through.

 

You could try adding. 250 ml of redex petrol treatment to your fuel tank and take the car for a good run and before the mot. My mot station advised me to do this years ago , and touch wood never had a car fail an mot in the last 8 years. I have found that it works better when you have about a quarter a tank of fuel .