Failed MOT high CO?

[Del]

Unfortunately, whilst flying through the MOT on everything else yesterday, my 1.9 failed on CO emissions. The MOT inspector’s opinion was a Cat issue. This is particularly annoying as I had a new Longlife stainless steel exhaust and Cat fitted less than a year ago – the Cat still being under warranty.

I have been fairly fussy with maintenance and servicing and the car seems to run smoothly, starts up well and is throwing no error codes. I am currently having an email exchange with Longlife about a possible warranty claim.

From what I have read high CO is otherwise caused by an overly rich mixture, has anyone experienced something similar or have ideas as to other possible areas of investigation? My 15-year old Lambda sensor is on my “to do” list but has not been throwing any codes.

[Robert T]

Before you go chasing your own tail on this, was the CAT fully up to temperature before they tested it? Having watched mine being MOT'd a couple of times, the guy I take it to leaves the engine running whilst he checks other things on the ramp and then does the emissions test at the end. By how much did it fail? I normally get a printout of the emissions results with my "certificate".

Cheers R.

[TitanTim]

Always a good idea to take the car for a good run if possible before the MOT.

Tim.

[PCSAM]

my old 1.9 would fail on CO as well mate . take it for good run (5 to 10 miles ) in 3rd gear at (70) if you can that will get the cat nice and toasty for you and i bet it will sail through the emissions test get it tested as soon as after the 3 rd gear heat-up

[Del]

The MOT inspector did two checks for the MOT both at oil temp 67C (VOSA min 60C)

Fast idle
First CO = 0.696% (target 0.3%)
Second CO = 0.638% (target 0.3%)

Idle
CO = 0.677% (target 0.5%)

After an initial chat with Longlife exhausts they suggested it was tested whilst a bit on the cool side and so I went to a “run” in the car and KwikFit agreed to quickly retest and the readings at oil temp 85C were:-

Fast idle
First CO = 0.666% (target 0.3%)
Second CO = 0.715% (target 0.3%)

Idle
CO = 0.565% (target 0.5%)

[BladeRunner919]

I wonder if it's worth trying a Terraclean? http://www.terraclean.co.uk

I've not used it personally, so I can't comment, but on Wheeler Dealers they had a Jag that was well over on emissions and the Terraclean did reduce the emissions considerably.

[Mike Fishwick]

I would have expected the ECU to give a code for an 'implausible' lambda sensor reading if the mixture really was incorrect for the engine conditions - the ECU would then be trying to weaken the mixture to correct the situation.

I would not be surprised if the emission test was carried out on a less-than-hot engine, so a really hot engine may make all the difference. Out in France our cars are tested for emissions as the very first step, the car being brought in at the time arranged specifically for this purpose - there is no 'Leave it and we'll get round to it sometime.'

But - if the problem really is rich mixture or incomplete combustion, either will produce excess CO in the exhaust.

I would look at simple things first - such as new plugs if they have not been replaced in living memory, and making sure that the plug insulators and plug caps are clean - rub them with WD40 on kitchen towelling until the towelling stays clean.

Likewise, clean the coils and the the HT cables, then check them in the dark - faulty insuation can provide pretty firework displays.

Is the air filter element clean? At low engine speeds the Motronic ECU does not use the airflow sensor signal, and assumes that the airflow matches the throttle opening, engine speed, and engine temperature. An obstructed filter will therefore make the mixture rich.

High fuel pressure would richen the mixture, and could be caused by a leak in the vacuum pipe from the fuel pressure regulator - or the pipe could have fallen off.

Dirty injector nozzles are a possibility - but not common - as this would cause a poor spray pattern, resulting in incomplete combustion. You could pull them out (replace their sealing 'O' rings) eyeball them, and make sure the nozzles are clean. A injector cleaning product in the fuel would also be an idea.

[Del]

Thanks for the responses.
I have ordered a new direct fit Bosch lambda - as I said earlier the existing one is 15-years old and is going to be reacting less quickly. Being a pre-OBD II car, my 1998 z3 only has one pre-cat lambda - annoyingly if it had a post-Cat sensor that would might help better pinpoint a specific Cat issue.
The car is running fine, doesn't burn oil and the current oil has been in since last November and still looks almost like honey. Running temperature is fine after fitting a new thermostat soon after getting the car in 2011. Virtually all vacuum pipework renewed. Current plugs have done 10k - they're cheap enough so I'll change those. The Bosch air filter is new only having done 1K. Will try some petrol/additive injector cleaner.

[Del]

Quick update:

Currently working on the basis that, my not quite 1-year old CAT, is not the problem.

1.Changed spark plugs – no noticeable performance difference.

2.Fitted my spare 2nd hand MAF – made quite a noticeable difference, the car revved more smoothly and seemed to have more power when going up hills in higher gears. Before the car did not feel comfortable below 2000 rpm in 4th gear and accelerating from that situation. Idle also seems smoother and quieter.

3.Replaced the single, pre-cat Bosch lambda sensor – my old one was stamped with a manufactured date of 12/1997! I took the view that after 16-years and 83,000 miles it would do no harm to replace it.

Car now runs much better. I now suspect that the high CO issue was mainly down to the MAF. I have done a bit of research and it seems that they can gradually “fade” without immediately setting an error code. It seems that older/contaminated MAFs can “over-report” airflow at idle (ie MOT testing conditions fast & slow idle) resulting in a slightly rich mixture (and higher CO) and “under-report” airflow under load causing a leaner mixture.

Will provide an update after MOT re-test later this week – fingers crossed it is sorted.

[Gazza]

Fingers crossed - thanks for the update

[Deano1712]

Hope its ok now. I think its unlikely to be the CAT. They don't do anything at idle, only fast idle when the CAT gets hot enough to do its stuff. Your readings indicate its running rich at idle. That could be the MAF. Its worth check for air leaks in the induction or exhaust too.

[Del]

Thanks for the helpful comments Deano - much appreciated
I am beginning to suspect the blame heaped on the new cat was just a flippant comment from a busy MOT tester. I've tested the replacement MAF on an OBD II live scan and at normal idle it is reading around 3.5 gps (grams per second) air flow which is about right for a normal 1.9 engine. I didn't measure the old faulty one before whipping it off but I would wager it was reporting an incorrect, higher figure - which I have read is a typical feature of an old contaminated one which is being insulated by a lot of fused-on dirt and contaminants.

[t-tony]

You might be surprised at how many cars fail the M.O.T test on emissions because they are NOT tested correctly,this is one of my biggest annoyances with VOSA,they know who these stations are,and all stations have to keep paper copies of test results for 3 months,yet they NEVER check these. Also when they come to review a recent test to make sure the test result is good - or bad they NEVER retest the emissions.WHY? IF a car fails the BET test(Basic Emissions Test) it should then be given a FULL cat test which should involve running the engine until the minimum oil temperature is achieved before testing the emissions again.Usually this will result in a pass.

[Del]

Quick update

Failed MOT again on just high CO. Had a lengthy chat with the tester who said that everything else, including the fuel/air ratios, where almost perfect. In his expert opinion when everything else is perfect and the heated-up Cat is still failing to clean the CO - it can only be the CAt at fault

Warranty claim for the Cat initiated this morning

[Mugs]

I had a longlife 'high flow' SS cat fitted to one of my previous cars and it only just scraped through the MOT by the skin of it's teeth so I ditched it in favour of a standard replacement cat and it passed with flying colours.
I like their SS custom systems but I have to say their replacement cats leave a lot to be desired.

[Mike]

My 2.8 used to have problems with MOT emmisions. Having moved to France and tightened a hidden screw at the back of the head it flys through French CT emmisions now. I have a powerflow cat back exhaust and the cat is only just OK in the UK tests.

[Del]

The Cat was faulty and has been replaced for free under warranty and the Longlife garage that fitted it, carried out a MOT f.o.c. and it passed with negligible Co readings which were well under the allowed level. In the end I was told the tale-tale sign on their equipment was the fact that my single pre-cat lambda (which is new) was fairly inactive and was not frantically trying to adjust an incorrect mixture being produced further up-stream. The fault therefore, had to be the Cat.

Longlife say this sort of Cat failure within a year is very rare and can’t understand it.

[t-tony]

After 2001 vehicle aftermarket cats have to be type approved but vehicles manufactured before this can be fitted with any old cat which will make the vehicle pass the emission test. Most of these are very small and are barely up to the job,some need to be replaced year on year. Beware of what you buy.