Failed MoT on Emissions

Hello All, my 2004 996 GT3 has just failed its MoT on emission big time ! and I would like some help on where to start looking. As you can see CO is way over the limit, O2 just outside the limit and HC also outside the limit. A bit of recent car history, the car has been off the road for the last 9 months while all the suspension, brake lines and cooling system was renewed and the under side of the car painted/treated. The engine/gearbox was also out of the car where coolant lines were replaced, oil coolers replaced, new clutch, replacement front and rear main seals changed, new plugs and coils, engine mounts. The various change-over/cycle were also replaced from the air injection system. The car seems to be running fine (although I've not driven it any distance as No MoT/Tax) and no faults coming up on the Durametric. What should I be checking for ? and what readings should I be seeing. As a last resort I might have to take it to the OPC but I would really like to fix the problem myself. thanks, any help/guidance much appriciated.

Is the exhaust system standard?

Here's a clip of the O2 readings, Bank1

And Bank2

The car has a Kline exhaust with some 200 cell Cats welded in place - thats how the car came - I have the paperwork the exhaust.

High CO and HC suggests cats, but low lambda suggest fuelling issue (running rich) so cats may be fine. I can't get the O2 sensor links to open. Does the GT3 have a vacuum regulator on the rail? If, so has it been connected properly? Is the MAF connected properly?

I'll check the MAF tomorrow evening. I'm not sure about a vacuum regulator, there's a fuel pressure regulator - i'll have a look at the diagrams.

Here's some screen shots of the O2 readings

Some ramblings below! The lambda of 0.9 is saying the emissions are very rich, the very high CO level is saying the emissions are very rich too. In your vids both the secondary O2's have high voltage which is also saying your rich. What is strange is your pre cat O2's appear to be cycling normally, that doesn't make sense! Seeing as you have a Durametric it would be good to know your maf voltage and kg/hr readings with the ignition on and the engine off (should be 1.0v), and also while idling (should be 1.25v ish) Also your fuel trims rcat and frau. Also see if any misfires are being logged. If the line connecting the inlet manifold to the fuel pressure regulator is disconnected that causes a rich condition, but you'd normally see that on the pre cat O2's, a fuel pressure check would rule that out. The way mot machines measure lambda is not affected by if the cats are working or not, so you definitely have a rich tail pipe. Worth checking the exhaust for leaks.

Can you take it on the long route to the retest for a good old fashioned “Italian tune up”? I.E. Get it warm then rag the nuts off it.

Thanks Mac, I'll check the MAF and fuel trim readings tonight. I'll also check the exhaust for leaks as its been off a few times. Benny, when I do eventually take it back for its retest I'll take the scenic route, I'll also add in some Cataclean to the fuel to help the cats.

Further thoughts, as you've had work done to the change over valves it would be worth checking they are plumbed in correctly, in particular the EVAP purge valve from the carbon canister Don't know if possible but check O2's are not connected back to front, or are crossed side to side!

Here are the MAF and Adaption readings

I also checked the exhaust for leaks, nothing that I could find. Also no misfires logged or any faults. I checked the lambda wires (blue lead/black plug to pre-CATs. and black lead/grey plug to post-CAT - can't be mixed up). The post-CAT leads are not crossed over.

Dodgy lambdas? Test them.

More ramblings Maf voltage with ignition on looks good at 1.01v, with the engine running maybe a touch high at 1.4v but if it was still warming up it could be that. Might be worth searching to see if 1.4v is normal for a GT3! Both your fra adaptions are 1 which is perfect ( has range of 0.7-1.3) this is the adaption that covers cruise. It may well be that you've not driven it enough for them to move yet. Both your rkats are negative, this means its pulling fuel at idle. It puts the engine management light on at -4.5% (and +4.5% in the other direction) so bank 2 is close. I was getting the management light on my Turbo due to the rkat adaptions reaching -4.5%. My maf voltage was 1.1 with the ignition on and around 1.35v at warm tickover. I replaced the maf and the adaptions changed to -0.5% and the maf voltage dropped to 1.0v with ignition on, and 1.25v warm, and I've not had the light since. So what does this mean for your car, well although your adaptions are negative they are still within the normal range, so it may just be that your maf is over reading a bit. Do you by chance know another GT3 owner whose maf you could try? Also given the fact your pre cat O2s are switching between the correct voltages, that looks normal too. What I don't know is how fast they switch on a GT3, a quick search says most cars switch between 1 and 2 times a second, yours is much slower in the vids at around 5 seconds, don't know if that is telling us your sensors are lazy, but then it's unlikely both would go wrong at the same time. Also you said it appears to be driving normally. There doesn't seem to be much wrong, I'd still want to prove the plumbing of the change over valves are correct, and the vac line to the fuel pressure regulator is on properly, and that there's no fuel in it (if the diaphragm has split fuel can come from there) . I'd also want to disconnect and temporarily plug the line to the evap solenoid to make sure its not constantly sucking fuel vapours from the tank/ carbon filter. I think I'd also get another garage to confirm they get the same readings. If you could, driving the car to get the lambda red hot in case they've somehow got coated in something may make a difference. The only thing which seems wrong on your car other than the emission test result, is the post cat O2s are reading high voltage in your vid. That does match with the emissions results, but in one of the other pics the voltage has dropped!

Thanks Mac for all your time and helpful suggestions. I'll check the vac line to the fuel press reg and check the fuel press reg. I'll run it for a while to get the cats hot and see what the Maf voltage reading is with it idling. I'll need to get hold of a camera to check the routing of the change-over piping.

I checked the vac line to the fuel press reg, was attached ok and also to the resonance tube. I didn't disconnect it to see if there was any fuel in it. I've dropped a note to my local OPC to see what the GT3 MAF should be reading at idle. I ran the car until it was fully up to temp and rechecked the O2 and MAF readings - still the same. Next I suppose I'll need to check the EVAP solenoid and try plugging the line.

My c2 fails with the 200cel cats unless you get them very hot, by all accounts its relatively hard to get them to pass with the OE ones when they are relatively cold as well (according to the specialist that i take mine to) If you're located anywhere near the North East I can recommend where I take it, they are very finicky and well up to speed with the foibles of testing 911's!

Question, could this bleeder valve cause rich fueling ? 99611012990 for the Porsche 996 GT3 . It is used in the fuel system, specifically for the fuel filter, fuel pump, and fuel accumulator . This part is essential for managing fuel flow and ensuring proper operation of the fuel system in the Porsche 99 6.

That looks like a solenoid attached to a bit of pipe. While not being familiar with GT3 plumbing I'm guessing it's the solenoid for the evap system. It should be easy to test. With no volts you shouldn't be able to blow through it, with 12v on the connectors you should!

I'll give it a try tonight after work.

OK so todays update, I removed the bleeder valve/EVAP solenoid, no power and the valve is closed, however if you blow through the tube really hard you do get a small amount of air passing, not much and a lot of resistance. When you put 12v on the valve it opens and its really easy to blow through (no resistance). I then plugged off the hose from the valve (isolating the bleeder valve) and ran the engine - no difference to the O2 or MAF readings. So there doesn't seem to be an issue with the EVAP system. I found this next step from a suggestion from deMort back in 2020 thread - I then disconnected the MAF lead when engine running, then ( clear fault codes ) .. this will reset adaptions .. then we can see what effect this has on the Rkat readings ... we are basically removing the evap from the system . Rkat as i'm sure you know is adaptions at idle .. with a small issue this will increase the quickest .. again only when the car is idling . frau and Frao are mid range fuel adaptations and flat out / high rpm adaptations .. these would show a larger problem with fueling .The whole point of adaptations is to try and maintain lambda .. the correct fuel / air ratio .. we can use these figures to try and see rich / lean and when the car is having to alter the injector duration to do this . Maf running and idle .. 1.35 is on the high side .. acceptable but a little high. So I did the above and after leaving the car to run for a while, the MAF was still reading +1.4v and the O2 readings were still 0-0.75v and 0.7v. The adaptations were -0.42% and -0.75% initially then crept back up to -1.5%/-3.0% and were continuing to increase. So next step, I've ordered a new MAF sensor to see if that makes a difference. We can rule out the EVAP system as plugging it off made no difference - as bleeder valve was only just passing with hard blowing I think I can leave it for now - let me know if I should change it ?.

Good work, few more thoughts What should a normally working car be doing at idle, firstly it reads the maf, coolant temp, and inlet air temp, then applies the relevant duty cycle to the injectors, which in theory should give the perfect mixture, then the o2 sensors check the exhaust stream , the gt3 has narrow band sensors, they don't measure the mixture exactly, but do swing from above stoic to below stioc. Low voltage to high!. The ecu adjusts the injector duty cycle to make the mixture swing back and forth across this point. If something is reading out slightly, say the maf, then the ecu has to add or subtract from the theoretical value. It helpfully tells you how much it's adding or subtracting in the form of adaptions. It won't do this forever and has limits before it puts on the engine light. Most owners haven't a clue where their adaptions are running. This means the correct gas mix is going to the cats, which can then do it's job and you end up with low co, low hc and lambda of 1 measured by the mot machine. On your car there is a contradiction. The o2 sensors are switching which means they should be sending the correct mix to the cats. The fact your rkat fuel trims are negative in theory isn't a problem as the ecu is still happy and you haven't got to the point it will put the lamp on, it just means something is not reading optimally. Half the GT3s may be like this (if the owners ever drove them lol) So heres the contradiction, if your o2 sensors are working correctly they are sending the correct mixture to the cats, then because you have aftermarket cats they may not work as well as std cats but you should still get a pass but with elevated co and hc, but the lambda as read by the mot machine should still be 1.0. That's the important bit, the lambda on the mot machine should still be 1.0, in fact if the cats had been removed completely you should still have a lambda of 1.0. The mot machine measures all the gases to work out the lambda and it should be the same as what the engine management is saying. The mot machine is saying your car is running at a lambda of 0.9 which is very rich. So something strange is going on, either your o2 sensors are wrong, or the mot machine is wrong or your exhaust is leaking, that's unlikely as that normally causes the lambda to go above 1.0, not below like your car. Thats why I said I'd like to see the results from a different mot machine. If the results are the same then it's definitely your car that has the problem. If it's your o2 sensors that are miss reading then that's strange too, if they are lazy and not switching fast enough the ecu is smart enough to detect that and put on the light. Also it would be unusual for both to go wrong at the same time. When I've seen o2 problems on other cars they've always been lazy and that's put on the light. I don't know how many miles are on your car but if it was my car, I'd get it checked on a different mot machine and if the results were the same I'd replace the pre cat o2 sensors . I'd also be doing some more research to see if I can find out how fast the o2s should be switching at idle. I also consider the maf to be a service part too, my turbo has required two, and is not far off needing another at 70k. Alternatively you could put it into a garage and hope they have the skills to not just parts swop until it works!

Something else to check, I've assumed above that as your o2 are switching the engine management thinks the lambda is 1.0, you can see the values in durametric, it will say "lambda bank 1" and "lambda bank 2" or similar. It will be good to know what the ecu thinks it is! Another question has your car been mapped, if it has some of the usual cat alarms could be turned off?! The ecu has all kind of tests it does to check both the pre and post cat o2's so it is strange that your not getting any

Hi Mac thank you for your very detailed explanations, it makes a huge difference to my limited understanding and I’m sure to a lot of other readers out there as well. The car is just coming up to 60k, probably classed as very high mileage !. Looking at the MAF sensor last night it looks like the original, it was certainly the early part number which has now been superseded. There’s nothing in the history to say if it’s been mapped, but it was set up by the previous owner for track use so maybe ?. Any previous owners of LJ04EDX I’ll check the lambda bank 1/2 readings. If the new MAF sensor doesn’t help then I may just change the two pre-cat O2 sensors. I’ll see about taking it to another Garage to check the readings.

Hi Mac thank you for your very detailed explanations, it makes a huge difference to my limited understanding and I’m sure to a lot of other readers out there as well. The car is just coming up to 60k, probably classed as very high mileage !. Looking at the MAF sensor last night it looks like the original, it was certainly the early part number which has now been superseded. There’s nothing in the history to say if it’s been mapped, but it was set up by the previous owner for track use so maybe ?. Anyone know I’ll check the lambda bank 1/2 readings. If the new MAF sensor doesn’t help then I may just change the two pre-cat O2 sensors. Then I’ll see about taking it to another Garage to check the readings. Here are the lambda bank readings.

Here's my latest update, I fitted a new MAF sensor, unfortunately the MAF and O2 readings were the same.

That 1.03 for bank one is strange, the ecu should be controlling it at close to 1.0, 1.03 is the limit for a mot pass, possibly that's saying the bank one o2 is wandering around and not following the actual o2 correctly! I take it that was with the engine fully up to temp as the sensors don't work until they are hot, they do have internal heaters to help so another thing possible is the heater is not working, that would normally result in a fault code though. What I'd do from here, I'd let it run for a 1/2 hr or go for a drive if possible to make absolutely sure everything is hot and look at the o2 sensing readings and rkat's to see if they've improved. Next I'd get it checked at a different garage to rule out the first mot stations machine. If that was the same I'd change the pre cat O2 sensors or get a good indy to change them. If new o2 sensors doesn't work I'd give up and send it to a specialist! If it does turn out there is something wrong with your car and the lambda on the second mot machine is still low I really don't get why you're not getting any codes!

Here's the next instalment of this saga, so having exhausted of trouble shooting options I arranged to take it back to my local garage to get an emissions reading on both banks. I added two bottles of Cataclean (instructions said 1) but I was thinking worth a try in case I have sticking/passing injectors and it will help clean the cats etc. I then went for a 40min drive. The car had only previously been round the estate a few times along with lots of starts on the drive/in my garage and had been off the road for 10 months during the refresh. So back from my run I went straight to the testing garage and got the readings taken, as you can see below a completely different set of readings all well within the limits. Unfortunately I was too late to get the MoT done so it will be back in on Tues where it will be going for a run just prior to it being tested. Hopefully everything will still be ok. So I'm not sure what has changed to get all the correct readings ?. Mac thanks for all your help, I'll give a final update on Tues - all being well....

Its literally just from getting the cats properly hot so that they actually work at least you should just be able to enjoy it after Tuesday!

Quotes source post: 2830987

FYI the Kline cats if early versions may have failed. The very same thing happened to me a few years back (I have a full Kline system and manifolds.) My car also failed on emissions. Kline told me their early 200 cell cats were supplied by a different company they use now and are prone to failure after time. I replaced them with a new set from Kline and problem solved. Contact Kline (if a specialist can confirm failure of cars) and tell them what happened, they may do you a deal on replacements as they know about this issue.

Final update, and good news the car has passed its MoT at long last !. As I said earlier I'm not sure why I was initially getting such high CO and HC readings and no errors showing on the Durametric, the car seemed to be running fine, sensors seemed to be working. I think the combination of giving it a good run after +10 months (a few runs round the estate probably wasn't enough), the cataclean and possible the new MAF sensor. Isysmam - Looking at the cats they are not the original Kline ones, so they must have been changed at some point. Anyway they are working. Thanks Jer.

That's a relief, my feeling now is that the garage tested the car far too early, and probably before the ecu had switched to closed loop control, that would explain the very low lambda reading which is what differentiated your result from a usual cool performance cat failure as suspected by Scottos. Whatever there's a lot of good info in this thread which I'm sure will help others and there's still a bit of driving season left to enjoy it!