+1 on the water pump brands- OE or Stewart only.

Water pumps will tell you they are going bad LONG before they throw up and croak. As the bushing fails (no bearings in OE pumps and 90% of all replacement units- search "Stewart hi flow water pumps" for true bearing tech info http://www.stewartcomponents.com/ind...tegory&path=84) the shaft will have play allowing for seal failure. The play can be detected by grabing the fan blades and pushing back and forth towards the engine. The seal failure is obvious when you find your X5 pees on the garage/ parking space after driving, or you find green/blueish stains just below the casting and pulley.

In the case of many BMW engines with the older plastic part material, the trick is/was to replace everything in the cooling system at 60,000 miles as preventive repairs(rad, thermostat housings, overflow tanks, upper and lower rad hoses, W/P, belts, misc heater hoses, etc). Cooling system parts just fail under pressure due to age/mileage. The plastic Rads would just blow off the side tanks the first time you hit the highway. Water pumps drop off the early plastic impeller, and the fan would exit into the rad, fans blades explode, etc. All events would cook the engine before you could find a pull off.:wow: This is why everyone posts, REPLACE EVERYTHING when doing a cooling system update, on BMWs.

Do you need to? The simple answer is No, with the following legal disclaimer: Do you look under the hood of your vehicle and KNOW what to look for? If not, then EARLY Preventive Maintenance will be your best option to have a great reliable relationship with your X5. There have been many examples of high mileage cars with no issues. Others in stop and go traffic (is there any other kind?), throw up/out cooling system parts long before 100K.

Almost everything in our vehicles will tell you as they go bad. Except electrical ones. They just stop working one morning or drop dead in the middle of working...:p:

Now back to work....

Water pump failures on ALL Bm's are the most common regardless of what fails indise the pump. It WILL fail.

They can wipe out your engine and alternator if water cooled. It will fail while on traffic with no shoulder in the poring rain at night.

I paid $1,000 to have the dealer do it at 88k for just the precaution. Once installed at the dealer it has a lifetime warrantee and will also cover most residual damaged if fails.

In the M52/M54 engine, the WP may or may not give any warning before going south.
- If it leaks from the front seal, then yes, you will see coolant dripping.
- If the impeller flies apart at high rpm, then no, absolutely no warning at all, other than a temp gauge pegging all the way to the right in the red zone!

On the issue or brand: OE, is OK but watch for composite impeller, personally I stay away from BMW WP.
- Stewart is good but no better than HEPU, plus Stewart is overpriced.

I have had my HEPU water pump in my 1998 528i (M52) now for 8 years, zero problems. The Volvo people use HEPU all the time.

If people want the best bang for their bucks, stick to HEPU.

The above statement is true for I6 engine.

V8 is another animal.

That sounds like it failed right after installation. If so, that is a classic case of infant mortality. A new pump is simply more likely to fail than an old pump, due to the number of early hour failures we see. It isn't necessarily a design problem, as you suggest, but usually a manufacturing problem. People don't realize that when they attempt to do PMs and replace components that are running well, they are trading off risks. There is more risk of an immediate failure, in exchange for a likelihood that it will go longer (over the medium term) to failure.

Of course, your old pump may have failed and you had no choice. But for those considering a precautionary water pump replacement.....

PM practice has moved on a long way from simply replacing components based on time and mileage. Modern PM practice includes managing the consequences of failure. I have never changed my water pump (now at 100,000 km on the X3) because I know it will be more reliable than a new one. However, I am tuned to watching the temperature gauge and know what the results of running the engine hot are.

I am referring here to rotating components and mechanical wear. The expansion tank is a different issue: if there are signs that it has hardened over time then replacing it can be cheap insurance. But that hardening has little to do with total miles on the vehicle, it has more to do with the environment the vehicle was operated in and the nature of the thermal cycles it has experienced.

Checking Cam's reference of the HEPU pump, doth my eyes deceive me?

Made in Germany???

http://c1552172.r72.cf0.rackcdn.com/285133_x600.jpg

What is your rationale for that statement considering you would likely replace it with an BMW water pump that is the same as the old one? I think the risk of a manufacturing defect in the new pump is low compared to the risk of wear on the old one.

About a year later, not right after installation.

I have this one installed in a 528 now going on 5 years. No problems what so ever.
The OE W/P plastic impleller was hanging on with a thread when pulled out at 61K.

Okay, sorry if this is a little long.

That is the classical assumption that drove many preventative maintenance practices for decades. In some cases it is true. But in many cases, not true.

Here is a summary that I didn't write, but which describes the practices we used for years in very complete, professionally managed, long term equipment maintenance contracts.

I have attached two charts from this article, below. Most expect that failure modes look like Pattern A. But from one study:

(my emphasis)

So what can happen with changing components out when they are working fine? Look at the second picture. In the case of TMV, I was asking if the second failure was soon after the replacement, which would put it in the more likely phase. Turns out it wasn't, it went a year. But you have to get through that early phase before you get to a phase that is essentially the same probability as an older pump (depending on the slope of the line as time extends out). There is obviously a limit to how long it goes. But to say that all pumps fail at 100,000 is not supported by the data. That is just a convenient round number at which a number of people think changing the pump is a good idea. And some of those changes will result in a failure soon after, because a great number of components have higher infant mortality rates than mid-life failure rates.

We should also distinguish between raw data about component failure rates (represented in the charts) with the additional failures caused by interventions. Even when done correctly (right fluids, right torques, right part, no contamination) we will see failure rates like the raw data. But add in the risk of intervention, and early failures are much more likely. When we used to calculate heavy duty diesel engine life for replacements (typically every 15,000 hours for a 60,000 hour contract) we would on average not see the same 15k interval on the second, third, or fourth life. That is partly due to ancillary systems not being changed out at the same time and causing some subsequent failures, and partly due to the failures due to the intervention itself.

Since we are on water pumps, there is a classic failure mode for mechanically driven water pumps (more so prior to the adoption of belt tensioners). A shop would change the belts, which wore regularly. They would sometimes over tension them. The pump would fail soon after. And the customer would say, gee, what a drag, and I just got new belts too. Happened so often it was predictable. If the belts were tensioned correctly, the pumps ran a very long time.

Obviously, things wear out. But simply replacing components without real data on when they wear out, and the resultant consequences of failure, it not cost effective. I suggest that with BMW cooling systems we have little real data on when they wear out. Look at the urban myths around plastic impellers, which people all seem to know about but which never applied to the E53. But at the same time, we have a strong sense of the consequences being high when a component fails suddenly, and owners continue to operate the vehicle while it overheats. It happens too often. So the change outs are driven more by the consequences, not an informed decision about the likelihood.

I have posted this or similar articles before. This particular one (which contains the above quotes) is here for those interested:
Maintenance Management

Jeff

Yes, HEPU WP is made in Germany. Very quality stuff.
I have had it in my 1998 528i (same water pump as X5 3.0) since May 2006, Zero issues.

60K miles (100K kilometers) is still early.
Personally, I'd do cooling overhaul every 10y/90K or so.