S1 Elise 111s MOT emissions failure

[alan]

Hi all,

I took my Elise for an MOT last week and during the emissions test the coolant res burst, hence MOT failure. However, the inspector did say that the emissions were looking bad, and it would likely fail. I didn't have the knowledge at the time to ask what was looking bad with the emissions so i have no further info. oops!

I replaced the coolant res and temperature sender, and also checked that the cooling fan works. so fairly happy with that side of things.

The Cat and o2 sensor haven't been changed in 15 years at least, so i guessing it could be one,or both. 

These are the tests that i have done so far.

Blocked up the tail pipes and checked for exhaust leaks. There is a very slight leak on the join between the cat and the silencer, but this is to do with a bad joint rather than rust and i can probably fix with some gun gum. I can hardly feel this leak when the tail pipes aren't blocked up.

Measured the voltage from the O2 sensor. From various websites i found the black wire was signal and grey sensor ground - does this sound about right? I was getting a steadyish 0.8v while idling. When i gave some throttle and released, i noticed the reading went down very briefly to maybe 0v when i released the throttle. I read somewhere that when idling i should see the voltage flicking from 0 to 0.8 twice a second, but i didn't see this, perhaps my multimeter isn't quick enough, or incorrect info.....

Any thoughts and ideas most welcome. If it is the Cat, does anyone have recommendations on where i can buy one?

 

Alan - Elise type 79 111S

[Lithopsian]

The O2 sensor voltage should fluctuate rapidly between values of maybe 0.2V and 0.9V.  You won't be able to see this with a multimeter, you'll need an oscilloscope or a fast data logger.  A steady value is a dead sensor and you will definitely fail an MOT.

Small leaks will tend to raise the lambda measured at the tail pipe even though there may not actually be anything wrong with the emissions.  Leaks before the O2 sensor will cause the car to run rich and probably fail on high HC and CO even though lambda may look reasonable.  The K-series engines are given quite a lot of leeway for lambda on the extended test so you can get through with some fairly noticeable leaks, but plug what you can to reduce your stress levels.

The OEM S1 catalysts are high quality and will last a long time if they're not abused.  Running rich will kill them as will too much vibration or hammering them for who-knows-what reason (eg. removing those pesky studs!).  A bad cat will shows as high HC and CO, but this can be caused by other reasons including various sensors on the engine being bad.  I wouldn't be replacing what is quite possibly a good and valuable cat without actually seeing an emissions test.  Lots of aftermarket cats available, most of them apparently quite lot quality.

Really need to see the test results to recommend fixes.

Had to give my failing memory a little while to come up with the OEM catalyst brand: Cheswick.  Try Landrover part code WAG104260 and you might be able to get a half-decent cat that will last for more than one MOT even if it costs a bit more than some of the eBay cheapies.  Lots of aftermarket offerings for MGFs, quality about in line with price

[alan]

Thanks Lithopsian for the detailed reply. Much appreciated!

I should be able to borrow a scope to look at the output from the O2 sensor, but not before the re-test is due. So my current plan is to seal up the minor exhaust leaks around the silencer and then cross my fingers:).

I have a gut feeling that it will fail, but at least I'll get the emission test results.

 

[Broadside]

Why did the coolant res burst???

[alan]

Hi, not quite sure why the coolant res went pop, but I replaced the res, res cap and ecu thermostat. Also i have confirmed that the rad fan comes on at about 102ish. I just put it down to old age!!

I've just had the re-test and as expected it failed on emissions. I haven't had the chance to read up on what all the numbers mean, but it looks like it failed by some margin!

Any feedback would be most appreciated!

[Lithopsian]

Yikes.  Running very rich.  About 10mpg rich?  CO is massively high and HC is fairly high (almost a pass, but a healthy engine will be in single figures) which means over-fueling, and probably not an issue like misfiring or a dying catalytic converter.  Lambda also shows a very rich mixture which is unlikely to be caused by anything downstream of the O2 sensor.  Does the exhaust smell pretty fruity even when it is warmed up?  Failed worse on the 2nd fast idle, which strangely enough suggests the catalyst itself is working.

Which temperature sender did you change?  The blue one for the Stack seems to be working OK, but the ECU uses the brown one.  The engine will run rich if it thinks it is still cold, but maybe not that rich.

Might be the O2 sensor, always a good suspect.  Not easy to check yourself.  If you can't get a scope, try one of the data loggers that you can run on your phone, it will tell you a while bunch of things like the O2 sensor voltage, fuel trim, etc.  You'll need a special cable though.  Somewhat costly and a a real pain to replace (on the OEM downpipe) just on the off-chance.  If you fancy unscrewing the O2 sensor just for practice and have a propane torch, you can run some very simple tests that will show if it is working.

Next on the list would be the MAP sensor.  Hopefully not the sensor itself because it is built in to the ECU, but the pipe that runs to it can split or even get detached.  You can check it very easily.  Look on the right side of the boot lip as you're standing behind the car.  You'll see a little black box cable-tied to a bracket, with two flexible hoses attached to it.  That is a liquid trap to stop fuel getting to the ECU.  Check that those hoses are firmly attached and aren't split.  Symptoms tend to include hesitation, erratic idle, or stalling.

Ironically, another thing that can cause the engine to run rich is the fuel pressure regulator or fuel pump itself starting to fail.  More directly, if the injectors themselves are getting lazy then they can over-fuel.  A really dirty air filter can cause similar problems.

Another common cause of running rich on the K-series is an exhaust leak before the O2 sensor.  It fools the ECU into thinking it is running lean, so the ECU pushes more fuel through.  Lambda at the tail pipe would usually be close to correct (1.00) in this case as the ECU provides just enough fuel to balance out the oxygen sneaking in through the leak.  A leak further down the exhaust will cause lambda to go high, so even if you have some small leaks they are nowhere near enough to counteract the over-fuelling.  Small leaks will often seal up when the exhaust gets hot.

After that, I'm all out of ideas