Hello there and welcome to the BMW Car Club of America.

If you are a BMW CCA member, please log in and introduce yourself in our Member Introductions section.

E90 Water Pump Dissection

Discussion in 'E90/E91/E92/E93 (2006-2011)' started by Zahnarzt, Jul 15, 2010.

    • Member


    Post Count: 61
    Likes Received:1
    Because I plan to keep and drive my 2006 330xi until the wheels fall off, I took the advice of those respected independent master mechanics and preventatively-replaced my water pump at 50,000 miles. I was on the fence as to whether or not I was going to do this, but I ended up getting a really good deal on the new pump, so I went for it. And what does one with advanced degrees in anatomical sciences do with the old one? Dissect it, of course! Almost as an afterthought, I decided to take some pictures.

    I was primarily interested in looking for points of potential failure. I've heard about problems with BMW water pumps, but I wanted to see firsthand. Taking this thing apart was not easy. From what I can tell, the unit is extremely solidly-made and not meant to be taken apart or serviced.

    Taking the casing apart presented no particular problem. Standard torque screws hold the housing together. The rubber seals inside seem very adequate. Once the cap comes off, the pump is divided into two layers. The top layer has the obvious impeller, and the other houses the motor. In the motor side, it's easy to see that none of the electronics are particularly intricate. In fact, everything is pretty macro. From an electronics perspective, the only real point of failure might be if one of the leads was mis-soldered during manufacturing and somehow shook lose over time. The leads on this pump were pretty tough, but I was able to pry them off with a screwdriver. The solder connections on the large central resistor were notably weaker. By weaker, I mean that I was able to pop the connections with a little effort. I'm not sure what sort of real world driving forces would be able to do this unless there were manufacturing defects involved.

    Removing the lead and resistor housing reveals the motor. This is where I can see actual potential for failure. The clearance between the spindle and the hard plastic housing is absolutely minute. A hard knock might push the spindle against the plastic housing for a split second, scraping and releasing a fragment that could lodge and cascade the release of many fragments leading to motor seizure. The rubber shock absorbers on the mounting points probably decrease the chances of this happening. The plastic involved is that really dense stuff that has little sheer strength. (i.e. If you dropped the housing from a foot above a hard surface, it would probably shatter or at least chip quite a bit.) I suspect this material may be water/coolant resistant, hence its application here.

    There isn't much to the impeller layer. The clearance between the impeller and its housing is significant to allow fluid to pass through. The impeller is made from the same plastic material, and I don't see how a metal one would help much. If something gets caught between the spindle and its housing, it's going to cause the thing to jam regardless. In such an event, I suppose you could argue that anything floating around wouldn't fracture more pieces off the impeller, but at that point, I'm not sure if matters.

    Suffice to say that because water/coolant is flowing through this thing at a steady clip, it might take a bit of a perfect storm for a fragment cascade to actually seize the motor.

    There wasn't much past the spindle layer other than the motor coils. Perhaps this is a more logical point of failure. Brushed or brushless, electric motors do fail, and this one runs all the time, so...

    Any other ideas as to why these things might fail?

    :D Rick
    • Member


    Post Count: 61
    Likes Received:1
    the rest of the pictures

    Additional pics attached.

    ForcedInduction guest

    Post Count: 358
    Likes Received:4
    The water pump is cylced on and off as needed but your point about high run time is valid. Your pump looks to be in good condition for 50K miles.

Share This Page