This is intended to help out all of those here who have not (yet..) taken apart one of our turbos and examined it. For those who have, please bear with me if im restating what might seem obvious.
Plenty of us have experienced blue smoke coming out the tailpipe as a result of the turbo leaking oil into the exhaust housing and the oil burning in the exhaust. Plenty of turbos have been replaced over the years because of this problem. I myself have replaced entire turbos to get rid of this problem, even when i didnt know for sure that the one i was putting on wouldn't do the exact same thing!
The natural thought is to blame the problem on a leaking seal on the exhaust side of the turbo. While this is always essentially true, it CAN be a case of oversimplication, and of regarding the symptom as the cause.
Now, I'm going to go through a basic pictorial explanation of how oil coking can cause your car to push oil out of the turbo and into the exhaust, something that is alluded to but not explicitly explained in the General Turbocharger FAQ in the Knowledge Center:
First, let's get familiar with the basic turbo center section. We need to be familiar with this because it will help understand the orientation/viewpoint of many of the following pictures.
This is what our stock log- and TII-style Garrett T3 turbos look like with the compressor housing and exhaust housings removed. The flange area which is at the top of the picture is where the oil feed enters the turbo. The flange area at the bottom of the picture is where the oil drains from the turbo into the drainback tube. The hole that we are staring straight into is the coolant return (exit), and on the side that you cant see is the coolant feed (entrance). If this turbo was mounted on a motor, this would be the view when looking at the backside of the motor.
This is a view of the top of the turbo, where the oil enters the turbo and your oil line/fitting threads in.
This is a view of the bottom of the turbo, where the oil drains from the turbo and where your oil drainback tube bolts on.
This is a view of the center section with the shaft and wheels removed and laid out. The view is the same as the 1st picture, with oil feed on top and drain on bottom.
This is a picture of the turbo shaft and turbine wheel (the exhaust-side wheel). There is a silver sealing ring in the middle of the picture. This is the 'exhaust seal' of the turbo. This is its uncompressed state. The gap in the sealing ring closes when the ring is 'squeezed' into the center section. When it is squeezed, spring tension keeps the sealing ring tight against the part of the center section that it rides inside of.
This is a view of the exhaust side of the center section, where that shaft and seal go into the center section. You can see a tiny groove in there.. the sealing ring rides just above that groove.
A slightly different view of the same area. From here, you can see the light gray journal bearing that the shaft spins on, and the darker gray snapring that retains the journal bearing. You can also see some of the oil feed holes in the journal bearing. This is where the oil comes from that leaks out of your turbo and burns in the exhaust. So, in between those oil feed holes and the tiny groove which is where the sealing ring rides, there is not a flat level surface for the oil to just saunter on over to the seal and leak past it. The oil oozing out from between the journal bearing the turbo shaft will fall into the area below the snapring in this picture, which is partially in shadow. Down there in that shadowed area, there is another hole that the oil will run through and drain down into the drainback tube.
This is a view up into the bottom of the turbo through the oil drain hole. The inside is being illuminated by a flashlight shining in through the area shown in the picture above. As you can see its pretty empty in there, but if the turbo shaft was in the center section you'd see it passing right through the middle of that illuminated area.
Another view through the same hole, up into the center section, from the bottom. What you're looking at is the backside of the journal bearing from 2 pictures up. You can see the snapring that holds the journal bearing on this side. Oil oozing out from between the journal bearing and the turbo shaft would fall straight down, towards the camera, and drain out.
I could not get the camera to focus at this angle. What you are seeing here is the same journal bearing as the previous picture, with light shining through it. Directly below it is a sliver of bright light. That light is coming through the hole i mentioned 3 pictures up. This is the only way i could photograph it. So, oil coming out the near side of the journal bearing would fall straight towards the camera. Oil coming out the far side of the journal bearing, which we see 3 pictures up, would fall down to where this smaller hole is, and then drain towards the camera.
This is a picture of me sticking a tiny pick up through the oil drain hole to get to that tiny hole from the picture above.
This is a picture from the outside, showing the very tip of that pick poking through the hole.
So, as you can tell, im spending a lot of time here describing that tiny hole, where it is, and what oil goes through it. Here's why:
ALL of this carbon came from THAT tiny hole. The hole was blocked COMPLETELY by it. What you're looking at USED to be oil, but it became this carbon mess from sitting in that tiny area after the engine was shut down and just cooking from the heat of the turbo. Essentially, when it just sits there the oil 'heat soaks' to an extreme degree and breaks down into this junk, which then clogs the hole.
In this picture from the Knowledge Center article i linked to above, you can see an example of the area of a turbo that becomes restricted in the upper right of the cut-away. However, that is not a stock Garrett turbo, and the way that one is laid out would make it much less likely for it to clog up in the same way this Garrett did..
So, lets look at this picture again. If the little oil return hole down in the shadows is completely blocked, what happens? The oil oozing from around the journal bearing will fill up the entire area between the journal bearing and sealing ring. When oil is draining normally, there is no pressure here, but when this chamber fills up, it will actually pressurize somewhat from the oil pressure feeding the turbo and being forced out of the journal bearings. Now, the sealing ring, like a cam seal or a rear main or lots of other kinds of shaft seals, is not really designed to hold back fluid under much pressure, so it leaks!
So, what is the moral of the story here? I bet, looking at all the pictures except the last one, that you thought this turbo was pretty damn clean. In reality, if i had tried to rebuild it (with a brand new sealing ring..) and reuse it without removing that carbon buildup, it almost certainly would have leaked oil on the exhaust side in spite of the new seal. So, obviously anyone rebuilding their own turbo should be absolutely certain to make sure that oil drain passage is clean and clear.
Also, this is some good evidence to support the idea that you should use the most heat-resistant (high quality? Not sure..) oil you can find, and change it on schedule. Common wisdom seems to be that synthetic oils hold up better to extreme heat. Also, if someone were to remove the coolant lines from their stock turbo (i have done this at one point due to some damage on the center section) or have a non-stock turbo setup that does not use water-cooling, it would be a good idea to let the engine idle and pump some cooler oil through the turbo to cool it down, or use a turbo timer, when shutting down the vehicle. I know that's stuff we've all heard before, and its really not the point of this thread to regurgitate it. BUT, for anyone who has never really understood WHY we are all told to consider that, you now can visually SEE why, and what can happen.