wenied wrote:
You shouldn't route it back to the engine! The filters collect water as well as oil.
I don't have a ProVent, but I do believe it can get water in the system, however if you take a look at the application guide at:
http://www.mann-hummel.com/industrialfi ... v_pid01=44
... you'll notice that the bottom fitting is identifed as an oil-return pipe.
Assuming that the bottom fitting contains water, I agree it should not be routed back to the engine, BUT the "return" from a filtering system shouldn't necessesarily contain water if it's implemented properly, which I think is where the problem may lie with many users.
Water is held in vapor at relatively low temperatures. The filter system's purpose is to filter out atomized oil, not steam/water vapor, granted it does contain a dessicant, but that probably gets spent relatively early in the life of the filter. Chances are that if the filtered air is leaving the chamber proper, it probably still contains the water vapor when it leaves. I'm guessing that since the provent has the outlet-to-air-intake towards the bottom side, that 90% of users end up having at least a slight upward slope to get the tubing to the air intake. If that path has a significant temperature difference, like with non-insulated tubing or PVC (and I'm guessing that using something like copper tubing inline would actually make this more significant), there's a good chance the water is simply condensing in the tube going to the intake and running back downstream to the ProVent.
I may be wrong, but think of it this way: If the water's coming out of the crankcase and getting trapped in the provent somehow, what would happen if you DIDN'T use the provent? It would go out the stock case-cover unit (admittedly along with some oil vapor) and go into the intake system (where trace amounts of water are common).
I'm not saying that I think anyone should connect a provent the way that most of the online examples show and then put the drainline back into the oil pan. I think that would be a bad idea. What I'm saying is that the design of OldNavy's filter (similar from a fluid-synamics perspective to the stock unit) is a better solution to the problem.