[Artmasterx]

[[[[ Update 6/22/2015 ]]]]
Everyone, I have compiled initial statistics based on 559 responses so far.

This gives the failure probability and the +/- 95% confidence intervals for that failure rate.

Everyone help out by voting!

Also, there will likely be a skew towards more failures given that many people with healthy water pumps may not check out the thread.


Plot of estimated failure rate during specific mileage intervals:



Plot of estimated failure rate before a specific mileage (perhaps more interesting to some of us):

(I didn't really propagate the uncertainty correctly, so the uncertainty is probably larger at higher mileages)

[Fundguy1]

Mine went to 199,968 miles.
2008 335i e93 mt6, Florida heat.

[paco89]

I am at 128k and still going. ***Bought the car at 119k**** so i have no clue if it was replaced before. Any way of telling?

[benton0311]

Mine started rattling at 78k miles so I replaced it. I drove it for another week or so after the rattling started and all went fine, I just kept a close eye on my coolant temps. I replaced with a new pump the following weekend.

I'm just happy I heard the rattling and the pump didn't just die on me at random. The sound was very subdued and you could only hear it after turning the engine off when the pump would cycle for a minute or so after shutdown.

[roadkillrob]

MY 2008 coolant flushed and replaced at 5 years old and about 15k miles, pump died at 19k miles and 7 years of age.

[Fitz335]

At 102k, Thermostat failed at 80k, been waiting for the water pump to die but she keeps on rolling. Have a new WP/Thermo in the trunk, waiting for some free time to change it out.

[jwalker87]

I wish I could revise my answer. When this came out I was failure-free at 70k or so. Then at 75k my water pump committed suicide 2 hours from home, with the replacement sitting on the bench. Argh!

[Welcome to NBA Jam]

This thread gives me a massive nerdgasm. OP I'll help you out tomorrow with some SAS output (if I have time)!

[Artmasterx]

Originally Posted by paco89

I am at 128k and still going. ***Bought the car at 119k**** so i have no clue if it was replaced before. Any way of telling?

If you were ever under the car and could see the date stamp on the part, that would probably be the best way. They may have also updated the part number, so that could be sufficient too. Not really sure though.

Originally Posted by jwalker87

I wish I could revise my answer. When this came out I was failure-free at 70k or so. Then at 75k my water pump committed suicide 2 hours from home, with the replacement sitting on the bench. Argh!

That's bound to happen with many of the responses, so no worries except for the fact it failed.

Originally Posted by Welcome to NBA Jam

This thread gives me a massive nerdgasm. OP I'll help you out tomorrow with some SAS output (if I have time)!

Glad to hear it, and feel free! I am a chemical engineer by training, and would hardly call myself a statistician. I am surprise the number of responses we have gotten, and the results have not change much as the last few hundred data points were added. I think my statistical analysis is generally correct, but the propagation of uncertainty probably not so much. Oh well, I think it hits the core of the matter.

Kind of interesting to see a bimodal-ish distribution in the time to failure. I am not really sure if that is a function of sampling bias or something physical with the pumps or possibly a bimodal type distribution of driving/usage behavior.

My car just had it thermostat replaced, but the waterpump is chugging along with 72k miles on it. My extended warranty covers me for another 3 yrs and until 100k, so hopefully it fails before then...

I bet BMW has extensive data on this; it would be interesting to see how close we are to the truth. I assume we will never know.

[pimple]

82k miles last month with sudden failure--noticed fan running high at a light; next light yellow warning; a minute later red warning at which point I quickly pulled over and shut it off. Operated entirely in a hot weather climate. 2007 335i. Expensive part to replace, but can be DIY'ed if one is so inclined/experienced.

[bahasad]

Car has hit 100.7K with original waterpump/t-stat. Fan seems to be running at a high rate more often than not, but coolant temps tend to stay in the 200-210F range. I made a 600 mile trip yesterday with long idle/park sessions, and car seemed ok.
If it doesn't shit the bed in the next 2000-5000 miles, I will replace it proactively.

[fazman]

2008 e90 335xi fbo, 105k miles (still original WP and TStat), wish i could revise my old answer.

[jgoens]

Mine died two days ago..26k

[Wilt]

Originally Posted by Artmasterx

[[[[ Update 6/22/2015 ]]]]
Everyone, I have compiled initial statistics based on 559 responses so far.

This gives the failure probability and the +/- 95% confidence intervals for that failure rate.

Everyone help out by voting!

Also, there will likely be a skew towards more failures given that many people with healthy water pumps may not check out the thread.


Plot of estimated failure rate during specific mileage intervals:



Plot of estimated failure rate before a specific mileage (perhaps more interesting to some of us):

(I didn't really propagate the uncertainty correctly, so the uncertainty is probably larger at higher mileages)

Can you explain how, with 184 replacements out of 601 total respondents (as of July 9, 2015), only 30.6% of all respondents HAVE replaced their water pump, but your graph of probability of failure is up around 90%? I am not well versed in statistics, but 90% Probability of failure does not seem to jive with less than 1/3 of ALL owners needing a water pump replacement!

[BeastMode335i]

just looking at that math equation up top caused me to have a cerebral infarction lol..

[Zooks527]

Originally Posted by Artmasterx

[[[[ Update 6/22/2015 ]]]]
Everyone, I have compiled initial statistics based on 559 responses so far.

This gives the failure probability and the +/- 95% confidence intervals for that failure rate.

Everyone help out by voting!

Also, there will likely be a skew towards more failures given that many people with healthy water pumps may not check out the thread.


Plot of estimated failure rate during specific mileage intervals:




Plot of estimated failure rate before a specific mileage (perhaps more interesting to some of us):

(I didn't really propagate the uncertainty correctly, so the uncertainty is probably larger at higher mileages)

I believe you're using the wrong divisors in most of your buckets, as they don't count cars that have made it to and past the bucket in question.

As an example, the divisor in the <60k bucket should be 610, not 243, as all of the cars in the higher groups passed through the 60000 miles group without failure (note this is using older number than currently reported by the poll, but the point holds).

If you do the math with all the cars that make it to higher mileage groups (both failed and passed) included in the "passed" numbers for the lower groups, you get the following results, sans confidence intervals.

[Fundguy1]

So average life us about 100k. And I got 200k. I'll take it.

[oregondonor]

Is the water pump failure predominately a N54, N55 issue? Quickly scrolling through the thread it seems most responses are from 335 owners.

[Fundguy1]

E9x. All types.

[hockeyplayer]

Add me to the list of failures. It happened last week in about 95 degree (F) heat here in NC.

Approximately 110,000 miles.

[Barefoot-335is]

Funny, I am at 55K miles and have replaced 2 Waterpumps by 44k miles...first one lasted 26K I think, the second 18K....just call me Lucky LOL

Edit: just found my original post in this thread, first one lasted 28k, which means the second lasted 16k miles, car is now at 55k so I got maybe another year and I will be doing another....Oh, joy!!!

[jt335i]

Bought mine new. 5 years, 8 months, 6 days, and 61,300+ miles since the first day I drove it home and it's still running good on the original pump. No signs of failing.