Oil leak from AOS (?) during track weekends

I have had my 996 since 2013, and I do some DE events with the car. During most track weekends I will find oil leaking off the front of the motor, and it appears the oil is coming from the AOS. I occasionally get some puffs of smoke out the tailpipes as well. I try to run with the oil on the low end of the dipstick, and if the oil level is ever higher, it seems that the symptoms are more pronounced. I do not get any of this during normal, street driving. Anyone else ever have this issue? Is this the sort of thing that the motorsports AOS is supposed to cure? Back around 2015 the original AOS was replaced with a new, standard AOS. That did not cure the problem.

Keeping the oil level low on the dip stick for a car used at track events is never a good idea, especially on a 996! After you replaced the AOS in 2015, the issue never got better? Did you find the previous AOS to be defective? Did you find oil anywhere where it shouldn't be?

Quotes source post: 15974143

At higher g's the AOS can become overwhelmed, and suck oil into the intake. This is more pronounced with the .2 cars ('02 and up) due to slight design changes. As you go faster with stickier tires, it can get worse (it is NOT good for your motor). So yes, this is why the Motorsport AOS is used in many track builds. Edit: as for the oil leak, that is not associated with Motorsport AOS or lack there of. The crankcase vent system may be compromised somewhere. The exact location of the oil leak should be determined.

Your symptoms are common. First. as said above, just make sure where your oil leak is originating. Unfortunately the AOS is one of the Achilles heals on a 996. Track driving puts a real strain on the AOS, and can show up on the track but not during normal street driving. However, it also means that the AOS diaphragm is getting weak and will eventually fail. Sudden failure of the AOS can result in a hydro-locked engine...…..very bad, especially if it happens on the track. I've just put in my 2nd AOS in 9,000 miles because of lots of track events. FYI - You can install the stock Porsche AOS for significantly less than the cost of doing the Porsche Motorsports version. The Porsche Motorsports version requires quite a bit of plumbing alterations if installed in a stock 996. Stock AOS is less than $150, the Motorsports version is around $800-$900 (plus all the extra installation labor). Lastly, the Motorsports AOS is not fail-proof either. There are instances of it failing too. Just budget $1000 and replace your AOS if that's the problem. P.S. For track events, you want to keep your oil level at the half-way point on the dipstick, i.e. not full. Check it manually on the stick, not with the display on start-up.

Thanks for the feedback! I installed a deep sump a few years ago, and I keep an eye on the oil pressure. No problems with that so far. I started my last track event with the oil level near the top of the dipstick (yes, I love having a real dipstick!), and blew some clouds of smoke. I wish I had a decent picture of the oil leak, but I don't. Looking at the motor from the front of the car on the driver's side, up where the AOS resides, there is oil on the AOS and dripping down the front of the motor below the AOS. Other than the mess, it doesn't seem to be causing any particular problem. I just keep an eye on the oil level. When I did replace the original AOS with a standard replacement, the symptoms returned the next time I was at the track.

Quotes source post: 15974143

This is one reason I'm glad I had my AOS changed out during my clutch service. I looked at the higher end Porsche motorsports AOS, but it was crazy expensive at around $800 compared to the $150 standard version. Since I didn't plan to track my car, I didn't give it a second thought, but I'm heard they are much better for tracking applications.

I bit the bullet and installed the Motorsport AOS. On sharp left sweepers I would get smoke, sometimes a lot. Never happened under any other condition. MS AOS fixed the issue.

I think I had this problem after my track day yesterday. After got home looked the car over and saw some oil had dripped down front of motor on passenger side. Following it up, I see some oil on this corrugated tube that runs under the intake manifold runners on the passenger side. The AOS itself is dry though. I started my tack day with oil slightly above half way on dip stick. Car has LN 2QT deep sump. The car ran strong, had solid oil pressure and ran cool for temps in mid 80s. Did not see any visible smoke and no one mentioned seeing any while I was on track. Since the AOS itself is dry, does that rule it out and I should be looking elsewhere ? Can has not leaked any oil during street driving.

Quotes source post: 16776537

The AOS line run on the back of the engine just next to the accessories. The black corrugated tub you see here I believe is the wire loom?

^^^thats not correct.... that black pipe is the line running from the AOS to bank 2. I�ll update this post with a picture. IMO if there is ANY sign of the AOS sucking up oil you need to test it ASAP. This is sign that you potentially dodged a bullet from hydrolock. Look on the marketplace, there is a user selling oil caps used for hooking up a manometer to test your AOS or make one yourself. I would also strongly advise you take a look at the UAOS solution from another user on this forum, especially if you track the car.

Quotes source post: 16776902

Well, your wrong. The AOS pipe is only corrugated close to the AOS on the left side of the engine. This is the right side and you can see the oil cooler, the AOS line does not run that close to it. Here is what the AOS line looks like: https://www.rosepassion.com/fr/schem...10714753-57258

I need to take some more stuff apart to identify origin. If AOS failing would oil likely end up in plenum pipes or the crossover behind plenum ? From location looks like it could be leaking from rear crossover connections then down corrugated tube. Again, AOS and connections to it visibly dry.

Fyi, my original aos line was corragated on both ends of the tube. I would guess that if the aos was dry, then the tube may be cracked. U may have sucked oil up through the aos and it drained out the tube. U should test the aos a precaution and check the tube.

Quotes source post: 16776921

Okay.... Well you're wrong.... See pictures below.... And one more point... you can see VERY CLEARLY the other wire loom you are speaking about below in his picture (see Red circle). Also see the additional picture below that one that CLEARLY shows the difference in what the wire loom piping vs the AOS piping looks like.

KWright, that’s definitely the right tube, I noticed the connection to the head clearly visible in your picture when I was tracing it last night. Now to figure out how it got coated in oil.

Quotes source post: 16776967

Yes, a failing AOS would cause oil to be pumped into the plenum. It is most likely leaking out from the rear crossover pipe as you suspected. Both L/R intake manifolds also have recessed areas where oil can collect on top, I would recommend cleaning these out if there is oil in them. I will see if I can find a picture and edit this post. Edit: I cant find my picture, but can take one when I run home for lunch. For now, here is the area where oil can pool up if the AOS is failing/sucked up oil.

Oh, I stand corrected. What year/engine was this type used?

Quotes source post: 16777041

MkII

Quotes source post: 16777030

I was able to get a good look at plenum and the connections last night and they are dry, or at least no leaking. I will have to get the plenum removed to get a good luck at the crossover, I will give that a go tonight. I have a new cap and manometer coming so I can test. I will clean all the oil I can find and put back together then test. Hoping the AOS ok for now and this was just a result of being over filled with oil.

Here is the area where the oil will be. These pictures are of bank two after removing the front crossover/throttle body pipe.

I puled off the TB and plenum. A thin coat of oil was present in the plenum, and the intake manifolds - No puddling. I was also able to feel around to the crossover pipe and same deal, light coating but no puddling. Makes me wonder if the hose I originally pictured may be split as the amount of oil that leaked down the motor seems excessive considering there was no puddling in any of the intake system, I am unsure of best path forward and would like opinions. I have manometer test stuff coming in tomorrow, so: 1.Clean and put everything back together - run the test and proceed. 2.Dig deeper into crossover and the corrugated pipe 3.Other Will the manometer testing reveal if the corrugated hose is split ? could it fool me into thinking the AOS is bad when its really the hose ? I really appreciate all the help so far, thank you guys a lot !

Your first picture looks like oil vapor that has been accumulating over time. More so than a oil gulping / AOS took a **** event. Did you see oil dripping from this area, how much oil was actually leaking? I would wipe it down, clean the exterior of the rear cross over and that AOS tube well. See if you can tighten up the clamps for the rear cross over tube at all. Put it back together and test with the manometer . I not sure if that hose was cracked how it would affect the manometer readings. But thinking this out; the AOS operates off a slight vacuum in the crankcase. When the AOS fails it causes the vacuum to increase, potentially causing it to suck up oil. If that hose had a crack it in, you would think it would have a decrease in vacuum. Now a decrease in vacuum (or should I say this hose sucking in unmetered air) should definitely change RKAT and eventually FRA values.

Quotes source post: 16776921

[Image-only post; no author-written text. Original image references:] https://rennlist.com/forums/images/smilies/rolleyes.gif

Quotes source post: 16779645

First time in my life I'm wrong and you found it, enjoy

Just tested AOS and came up with -4.10... Anyone know if this is within factory specs ? Got the car warmed up with a 20 minute drive and do not see any oil after cleaning up what leaked form the track day. It ended up being a very small amount of oil that did a good job spreading. I started off track day with 3/4 full and ended with a bar or 2 above half.

Had a repeat of the leaking and this time verified smoke out the tailpipe at startup and while on track. Spent a little bit more time contorting myself under the car to try to confirm the leak is AOS (I’m changing it anyways, want to make sure not multiple issues). Managed to find what looks to be the beginning of the leak under the coupler connecting the manifold crossover (resonance pipe ?) to bank 1. Looks like oil leaks from there, rides it’s way down the corrugated tube directly underneath the crossover until leaking at the lowest point above bank 2.

Pulled TB off and knew it was no bueno right away. Plenum had a thick coating of oil to the point it started to run when I got it out. Bank 2 intake manifold had puddling of oil Bank 1 had very little oil in it comparative to bank 2. The crossover pipe had what feels like a thick film like the plenum... I have not gotten it off because it must have last been installed when the engine was out prior to me buying the car....When the AOS was last replaced ironically enough and the bolts for the worm clamps are not visible 2X and will take some major finagling 2X. I am wondering if when I ran the AOS test previously, it was bad - my cracked duct taped oil fill tube allowed enough of a leak to make it appear to be functioning when it was not. I had a much lower - 1/2 QT lower oil level for this track day than the last when I had the same, although not as bad problem. I think - the oil in the plenums in the amount seen points only to AOS but would be welcome to hear thoughts from others... ETA... My smoke show on a hard left hander

Ouch! It's a good thing you caught that - a big slug of oil and you'd have had hydrolock city!

When I replaced my failed AOS with the UAOS the corrugated tube just fell apart, and it was the source of oil leaking in that area. So now new tube installed as well. That's a whole lot of oil in the TB! Which is where the "air" from the AOS dumps back into the "fuel system". I believe it's a closed system from the AOS to the TB so that would be the only source of oil. Some of the 996 guru's don't like the Motorsports AOS so change to the UAOS, much easier to repair when the AOS next fails, Plus there's a oil drain on the UAOS accumulator for easier servicing.

Quotes source posts: 16816642, 16816702

You caught it ! If it wasn't for you letting me know about the plume at startup I would may very well have found out the hard way - THANK YOU Aaron ! Yeah, I'm just not sure...will find out when everything comes apart for the UAOS. Is a replacement for the corrugated tube in the kit or do I have to buy separate ?

Quotes source post: 16816711

No, you have to purchase that separately. The hose that is included in the kit is the one that runs from the AOS to the Accumulator to the TB.

Quotes source post: 16816711

You�ll have to buy it separately.

Quotes source post: 16816740

Cool, good to know, thanks. Any other misc items you suggest ? I'm likely to be cutting it a little close between receipt of parts and next track day - I'd like to be as prepared as possible.

Buy one of these and replace the old brittle plastic one with a metal one: https://www.rockauto.com/en/moreinfo...8314&jsn=10413

I just replaced this corrugated tube on my 3.6 yesterday in fact. Mine had rubbed up against the coolant level sensor and rubbed through, causing a messy leak on the left. Possibly cracked on the RHS also. Happily my inlet plenums seemed dry compared with these photos, although I don't track the car.

Quotes source post: 16818472

Can you give rough overview of the install ? Ran a smoke test using a cigar, hand transfer pump and my AOS tester oil cap - could not identify any leaks...little bummed did not get a "smoking" gun...see what I did there haha. Seriously though - not the result I was looking for.

Here's the writeup I posted on 911uk in follow up to earlier discussions I'd had there... I got the new AOS pipe a week ago and have been flipping back and forth between changing it or not since then. Seems a painful job to do when you're going to pull the engine before too long anyway, but on the other hand I'd like to try to track down the source of my leaks so I know what I need to do when that happens. For example, will I need to pull the cam covers and reseal those? Anyway, the good news is my engine only seems to have leaks in three areas - the left, the middle and the right! The middle one is not so bad and is likely the rms which I think is original. On the right is a bit of a weep, has started to make a mess and drop a little oil on the garage floor in the 1000-off km since the last cleanup. The left is pretty bad and has historically coated the engine bay and heat shields, and leaves a decent dribble on the floor. Poor form. Today I said to hell with it and gave it a crack. I started by trying to disconnect the pipe from the AOS, that popped off without too much cursing so I went ahead and remove the throttle body and front (ie rear...?) inlet plenum. This revealed the.... in a vee engine it's the 'valley', but here I'd refer to it as the 'high plain'. The high plain was clean and tidy and had no oil. The AOS vent line was zip tied to the rear (ie front) inlet plenum - the one with the vacuum valve. I then set about removing the screws from the bank 2 connection which is obscured by a veritable rats nest of pipes and wires. Noticing the high current battery wires in that area I'd previously disconnected the battery (with ignition switch in position 1 thanks to the internet). Removing the bank 2 connection opened up a gaping hole directly over the camchain, this got carefully cleaned up and covered to prevent any debris falling in. The pipe then threaded out without too much complaint but did make some interesting aged-plastic cracking sounds as it came out. I oiled the orings and bagged the ends of the new pipe for reinstallation, then threaded it back through (typically, this takes 10 seconds to write, but about an hour to actually do when you're doing it for the first time yourself). Only the AOS-end is small enough to fit between the manifold runners so it has to go out and in that way. Reconnecting onto the AOS was fun, I tried a few directions but could not generate enough sideways force to click it home either from above or below. In the end I used a long screwdriver to create a fulcrum and push against the lip on the connector from the underside. A bit of jiggling and 'click', it was home. the bank 2 connection just required a bit of care not to drop anything in there and to fit the hard-to-access screws. Reassemble the rest of it, re-learn the window and throttle stops, clean up and take it for a lap of the block and all seems well so far. It was very brief but the engine felt and sounded smoother. I found a hole where the pipe had been rubbing against the coolant sensor - I hope that is the source of the major leak on the left side which was making such a mess. Perhaps there is micro-cracking which is causing other leaks on the right, but I am now also giving closer inspection to the cam sensor on the front of the head which seems wet and might have been being blown backwards. At least it's easy to access! Cheers

Here's the front of my bank2 prior to cleaning. You'll note the corrugated tube is dry between the manifolds, but it was wet on both sides. On the left there was certainly a hole there from rubbing on the coolant sensor, on the right I'm sort hoping it was actually the cam sensor leaking and getting blown upwards/backwards to create the mess over the injectors etc, then running down over the head. Could also be the cam bridge, or even the head gasket although the latter I understand is pretty unlikely (hope so!). I have cleaned this area now and will drive for a week and have another look, meanwhile get a new sensor oring and change it while I'm there.

Thanks for posting the write up - your picture looks pretty similar to mine... Although I caught mine right away so the build up on mine is far less. I may give replacement a go...

Quotes source post: 16818836

My car was new to me a few months ago and the leaks were a known issue that had been there a while. I'm gradually chipping away at resolving and cleaning up the mess from a few years with some degree of leakage in a few locations. Before too long will be an enigne out service to do the rms etc, cam bridges if they need it, and properly clean up the engine bay and replace the coolant tank. Maybe I should fit a UAOS while I've got it out, I certainly don't want to try to replace that damn aos in situ, the pipe was awkward enough. This is how it is presently, with the new aos pipe and the right bank area cleaned up to try and detect the source of the oily nile!

Zbomb, I just found another hole in mine where it had rubbed through one convolution against the inlet runner #5. Have a look on yours.

Quotes source post: 16819098

Will do - thanks. Ordering the tube tonight... Once it gets here I'll probably lay under the car for an hour or so looking at how much its going to suck before finally starting the project - thanks for the motivation.

Quotes source post: 16820861

except its done from the top so open the trunk, neil at the motor and stare at the top for an hour .

Well after the smoke test did not point me in the direction of the oil leak last weekend and since the weather is favorable today and I no longer have a cracked and duct taped oil fill tube to skew any results, I decided to take take the car out for a drive, get it warm and retest the AOS with my manometer. Last results were -4.2. Car blew out a puff of oil (residual from sitting since last track day - maybe) at startup but otherwise drove great. Tested the AOS when I got back to the house...not the results I was looking for, really wanted a conclusive - AOS bad result so I could be confident the UAOS will resolve all issues, along with changing the corrugated tube before my next track day on the 1st of SEP... Hoping I don't have something else wrong...