MOT emissions fail

[BackRow]

So the last part of getting my ex drift sigma caterham registered was to pass an MOT test.

Unfortunately it has failed (pretty badly). MOT results

The Mechanics think the Catalytic Converter is a high flow/ sports one... but as far as I can tell it's stock.

There was an issue where the immobiliser wasn't working properly which is now fixed, but as a part of that the mechanics had been trying to get the engine on all day. This makes me think there's unburnt fuel in the exhaust and possibly in the catalytic converter.

I was wondering if anyone else has had similar problems and/or knows what to do?

Looking at what emissions suggest online there are two possibilities it's not getting enough air, which I wouldn't know why. Or has unburnt fuel in the exhaust... which is what I suspect.

Any help or thoughts are greatly appreciated 

 

[simon_h]

Lambda sensor I would say. It's not even trying to go closed loop and get to lambda 1. 
Stick a new one in and see if that helps. 

[John Vine]

Clearly your mixture is well on the rich side.

I'm wondering whether the engine and cat were warm enough prior to the test.  "Temp gauge showed warm" presumably refers to coolant temperature?  Coolant should be around 80C.  How hot was the cat?  Did you drive a reasonable mileage (say 10 miles) immediately before the test?  This is important as the cat has to be really hot for a satisfactory test.

Also, your lambda sensor/wiring could be suspect, or maybe the coolant was simply not hot enough (60C) for the sensor to begin operating in closed loop.  You could try replacing the sensor, but they're not cheap.

The best way to investigate would be to run the Easimap diagnostic software (on a laptop) while hooked into the diagnostic port.  This would require the (expensive) MBE cable.  There are quite a few folks on here that can advise.

JV

 

[Jonathan Kay]

As above.

Plus.. are there any previous data?

Please could you add a photo of the exhaust system.

Jonathan

[BackRow]

I didn't think the cat was warm enough either, but the garage told me they had got it really hot, checked on the odometer and they'd only done 2 miles... My drive there doesn't really affect it as I left at 8am but they only tested the emissions at 4pm. (Due to the imobiliser issue).

For the retest I'm planning to start by going the wrong direction and getting onto a local bypass. In total it will be 19 miles. They are able to get straight to the emissions tester if the retest is done at the end of the day.

I'll note that It hasn't really been driven in the last year (was going to be a project for someone else to get it road legal). So the fuel has been sitting in the tank, I've put 22L of premium into the tank to hopefully mix with the old if that's the issue. (I'm aware that it normally has E10 but if your not driving it over winter it's wise to put in premium)

I think at this stage the retest with a long drive to the garage is the best option, hopefully it just wasn't warm enough.

 

If not I guess I'm almost certainly going to have to look at Easimap and the MBE cable

[BackRow]

No previous data as the car was was used by a drift company for about 7 or 8 years. So this is it's first MOT.

As for the photos I assume you just want to see the silencer and Cat?

Everything except the Cat I assume has been with the car for 9 years. The Cat I can't actually say, it either came with the car when it went from the drift company to the dealer or the dealer sourced it from another Caterham.

Ben

[Jonathan Kay]

[John Vine]

The 2-mile bit may not be relevant if the tester ran the engine until it was really warm.  At the very least, I'd expect him to check for hot coolant hoses or a running cooling fan.  "Temp gauge showed warm" is not necessarily the same thing as a minimum oil temp of 75C as it takes a lot longer to bring oil up to temp.

Your plan to take the car for a long run immediately prior to the test is a good one.  If it still fails, it's diagnostic time.

Good luck!

JV

[John Vine]

There's an interesting discussion here from the MGF/MGTF Owners' Forum on this very subject.  Although they're talking about K-series cars (and 11 years ago), the principle will apply equally to Sigmas and Duratecs. 

VOSA (now DVSA) state:

"If a vehicle does not meet the BET [basic Emissions Test] limits then a Full cat test using vehicle specific limits must be carried out and an oil temperature measuring device should be used. However the IM [inspection Manual] does make allowance for vehicles where it is not possible/practical to use one."

In other words, the tester should wherever possible use a probe to measure oil temps, and not simply assume that "warm" on a coolant temp gauge is sufficient.  The probe would be inserted down the dipstick tube.  I don't know how practical that would be on a Sigma.

They also say:

"Though not always possible it is advantageous to present the vehicle at operating temperature and close to the appointed time."

In other words, always do an "Italian tune" prior to the test.

JV

[56]

The only important thing to take from those results is the lambda values - it's running very rich for some reason.

Until the lambda is within range the cat won't be functioning anyway so you can ignore that for now.

Sorry I don't know the sigma engine very well but I would start with checking the lambda sensor, coolant temp sensor and throttle position sensor.

[John Vine]

Until the lambda is within range the cat won't be functioning anyway...

I'm not sure I understand that.  Could you elaborate?

Apart from a cold(ish) cat, my current suspicion (which a retest might well prove to be nonsense, of course) is that the coolant may not have been warm enough for the lambda sensor to begin operating in closed loop.  The 60C trigger point is quite sharp:

 

This graph is from my Duratec, logged during warm-up. I imagine a Sigma would be similar.

The top trace (pink) is coolant temp, the blue trace is lambda, and the horizontal black trace is target lambda (1.0).  The horizontal blue line to the left is at about 1.2.  Note that it took 5m10s for the coolant to reach 60C (on tick-over) with an ambient temp of 12C.  Of course, the OP's tester may well have accelerated warm-up by running the engine at, say, 2500-3000 rpm.

JV

 

[56]

This image is stolen from Umicore. It shows that a common three-way catalytic converter needs to be very close to lambda 1 for best efficiency of all 3 conversions. Below lambda 1 (running rich) the ability to reduce HC and CO drops off rapidly. It may still be reducing NOx levels but those aren't checked on the MOT.

[John Vine]

(Umicore linky)

Thanks for that -- very interesting.  I hadn't realised that the lambda range for efficient conversion was quite so narrow!

JV

[BackRow]

Just an update - I'm also replacing the Lambda sensor - Seen a few posts on it failing on Sigmas, Supersport in particular.
Considering mine is ex-drift it's quite likely this is the case.
I saw one post that mentioned they weren't aware it had failed until they failed the emissions.

However there's also a chance they didn't get the coolant up to temperature.

For the sake of not having to do a 3rd emissions test I feel the £68 for a new sensor is worth it (especially when considering mine would be about 9 years old and run with no cat).

I've ordered this.

[BackRow]

It passed the MOT.

They did a test run before doing the proper test - it passed this, while they got the paperwork to do the real test it was idle for about 5 minutes. In this period it had managed to cool the cat enough for the HC to read double over allowed.
Kept the engine at 3k revs for about 3 minutes before trying again to find the HC was now well below the limit.

I'm not sure if it was the new Lamda, but it may be worth others knowing that the cat can get cold really quick.

[Jonathan Kay]

Thanks for adding the outcome and those details about the timing.

Jonathan

[Golf Juliet Tango]

Good news.

[John Vine]

Good news. Thanks for the update.

JV

[StevehS3]

#15 just out of interest, did the engine seem different in any way after changing the lambda sensor? 

[John Vine]

Re #19:

That's a good question, given that there were two simultaneous variables: cat temp and lambda sensor.  I'd be very interested in the answer.

JV

[BackRow]

Good Question.

I'm not a mechanic and this is my first Caterham so I wouldn't exactly know.

However I didn't notice anything different when the new Lamda sensor was in- I'm 95% sure it was just getting the catalytic converter warm.

I don't regret buying a new one however as it was about £70. For peace of mind I felt like it was worth it.

[StevehS3]

Thank you. I would have done the same and replaced the lambda.

As John said, I was wondering whether it was the sensor or the cat not up to temp.

I would keep the other as a spare as it might be ok. My last car, a K series, had a faulty lambda sensor (heater failure in the sensor) and I could tell there was something wrong. The engine was a bit lumpy until it was fully up to temperature and it seemed like it might stall at junctions. 

[John Vine]

I would keep the other as a spare as it might be ok.

I agree.  First, it would be worth swapping it back in just to check.  Then, if it is ok, carry it in the car (plus a 22mm open-ended spanner!).

JV