91 captured repliesDiscussion pagination complete · Source accounts, not a confirmed diagnosis
THIS DISCUSSION / SOURCE-DERIVED OVERVIEW
Discussion summary
YOUR RESEARCH996 water pump failure with pics
Reported or suspected cause
original one
Reported cause
No successful repair reported for this finding.
1
Identified cause
0
Suspected cause
0
Reported fixes
Read evidence and context
Explicit resolution reports before identified causes, performed actions and suggestions. Within each level, matching component labels precede direct query mentions and inherited thread context; successful versus unsuccessful discussion counts, owner confirmations and independent discussion support determine order. This is not a probability of success.
Discussions that report identifying a cause
1
0 original-poster resolution reports. Counts can overlap; they are not unique people or a success rate.
0 discussions have direct query context; 0 report a fix in that context. Other evidence may inherit the discussion’s subject.
Reported cause
I just changed the one in my 04 Boxster as it developed a shaft seal leak and it was the original one.
Leading evidence: 'original one' (1 discussions report identifying this cause; no attributable repair resolution). Across 1 matching discussions, the most frequently mentioned possible cause is 'original one' (1 discussions; 1 report identifying it and 0 suggest it). The most frequently discussed action is 'Check level and antifreeze protection level' (1 discussions). For that action, 0 discussions report a resolution and 0 report an unsuccessful attempt. No action has an explicit, attributable resolution report under the current rules. These are forum reports and suggestions, not a confirmed diagnosis.
Source reports, not a diagnosis or a success probability. Counts refer to discussions and can overlap.
Explicit resolution reports before identified causes, performed actions and suggestions. Within each level, matching component labels precede direct query mentions and inherited thread context; successful versus unsuccessful discussion counts, owner confirmations and independent discussion support determine order. This is not a probability of success.
Discussions that performed work without a confirmed outcome
1
0 original-poster resolution reports. Counts can overlap; they are not unique people or a success rate.
0 discussions have direct query context; 0 report a fix in that context. Other evidence may inherit the discussion’s subject.
Performed · outcome unconfirmed
He also cleaned my air mass sensor and throttle body, took care of a vacuum leak, and checked the suspension.
Explicit resolution reports before identified causes, performed actions and suggestions. Within each level, matching component labels precede direct query mentions and inherited thread context; successful versus unsuccessful discussion counts, owner confirmations and independent discussion support determine order. This is not a probability of success.
Discussions that performed work without a confirmed outcome
1
0 original-poster resolution reports. Counts can overlap; they are not unique people or a success rate.
0 discussions have direct query context; 0 report a fix in that context. Other evidence may inherit the discussion’s subject.
Performed · outcome unconfirmed
Just checked the Warehouse33 site for the low temperature thermostat and the shipping fee for a $46 part is $20.
Explicit resolution reports before identified causes, performed actions and suggestions. Within each level, matching component labels precede direct query mentions and inherited thread context; successful versus unsuccessful discussion counts, owner confirmations and independent discussion support determine order. This is not a probability of success.
Discussions that performed work without a confirmed outcome
1
0 original-poster resolution reports. Counts can overlap; they are not unique people or a success rate.
0 discussions have direct query context; 0 report a fix in that context. Other evidence may inherit the discussion’s subject.
Performed · outcome unconfirmed
He also replaced the serpentine belt and oil change but I gave him the oil and filter.
Explicit resolution reports before identified causes, performed actions and suggestions. Within each level, matching component labels precede direct query mentions and inherited thread context; successful versus unsuccessful discussion counts, owner confirmations and independent discussion support determine order. This is not a probability of success.
Discussions that performed work without a confirmed outcome
1
0 original-poster resolution reports. Counts can overlap; they are not unique people or a success rate.
0 discussions have direct query context; 0 report a fix in that context. Other evidence may inherit the discussion’s subject.
0 original-poster resolution reports. Counts can overlap; they are not unique people or a success rate.
0 discussions have direct query context; 0 report a fix in that context. Other evidence may inherit the discussion’s subject.
Suggestion
When I had my Acura at the shop for the 2nd TB change 6 months ago, I spoke to their mechanics about what fails or what doesn't if you don't do the TB service.
Counts describe captured discussions, not unique people or a diagnosis. Author identities are not reliably stored. Exact duplicate evidence is counted once across matching discussions; one discussion counts at most once per action/state. Multiple repairs reported together remain a combination. Suggestions and failed attempts do not establish a fix. Matches cover whole discussions and can include related faults, model variants and incomplete threads.
91 eligible posts · 0 unverified posts · 1 discussions match the subject in their title or opening post.
Rennlist: 1 discussions
0 evidence records excluded for query context · 0 excluded by outcome qualification · 0 adjusted by outcome qualification.
1 acknowledgements have no verified action link; they do not establish a repair result.
Outcome without a verified action link
I have tried back flushing, and even made a device to do this using pressurized shop air and water, but it didn't work.
Water pump failed, plastic impeller disintegrated, and the face of the block seems to be scored. All I've read in these forums suggests that if I throw a new pump in it, it's a ticking time bomb. There could be plastic blocking some critical passages, and the block surface may be too badly scored to build proper pressure.
The only alternative appears to be a full rebuild. What do you think? Are my fears overblown? Is it only a small percentage of engines that suffer catastrophic failure after a water pump failure like this, or a very significant number?
You can remove the thermostat and see if you can find any broken pieces there. The rest you can try to (reverse?) flush out. I think a few have done that here with success.
I already had the thermostat out since I was going to replace it at the same time. Nothing there, completely clean.
My wife was driving it at the time. As she pulled up, it was billowing coolant smoke that was blown onto the exhaust. I read the temp gauge right away and it was about 2/3, definitely way below red, so initially I wasn't too worried. Still some coolant in the coolant tank as well, but bottom line, the plastic could have circulated all over.
May not be catastrophic. Pull the thermostat housing, fill and flush, rinse and repeat.
You may be able to get the debris removed, but be patient and diligent for rogue plastic.
+1 on the reverse flush but you'll need to do some puzzle building to see if you got all the pieces. If it were me I'd try high pressure if possible.
Curious: How many miles in the clock?
It went 1 mile too far. This is why preventative pump replacement is worth its weight in gold. This has NOTHING to do with mileage, the WP is in service 24/7-365, as it is immersed in coolant.
You must account for all broken impeller pieces, else you will have a great chance of future silent issues occurring.
These pieces leave the WP and then travel into the bank 2 cylinder jackets and head first, the worst place they can go.
Jake.....It would be very helpful if you could point the poster to a DIY project on how to get all the pieces out of the system. ie:- how to back flush, if this is needed, or other methods you would use to get to the missing pieces of plastic impeller. Thanks in advance.
I can't because a back flush will do little to nothing to help this. The pieces don't all break off at once, and with coolant still flowing the pieces are pushed through the system by the remaining impeller blades.
I have tried back flushing, and even made a device to do this using pressurized shop air and water, but it didn't work. Where the pieces travel inside the system as soon as they leave the water pump is the issue.
In my classes I instruct techs NOT to bill for this service, because if they are unsuccessful and even one impeller blade is left in place, and finds the wrong spot in the head to be jammed, leading to a localized hot spot and a resultant crack, that it will be their fault.
in every instance that I have seen the heads had to come off to retrieve all the pieces. In some cases the pieces made it back through the entire system and traveled through the replacement pump, thus taking out its impeller blades, too. That's a rare one, but when this many blades are missing, it can happen.
No easy answer or action for this case. You might retrieve some blades but I doubt you'll get them all. We typically utilize the ultrasonic cleaner with the head immersed in it for 20 minutes to knock the pieces loose, then shop air will blow them out. Nothing else can get them out once they are deep in the head and jammed there, even with the head off the engine.
Every single piece of my failed water pump was sitting right down in the thermostat housing area. I was able to take the thermostat off, get all the pieces just by reaching in there and rebuild the entire impeller on my garage floor. When I changed out my pump, I did not do a flush, I just refilled and burped the system to get any air out. When I changed the coolant overflow tank a few months later I did the coolant flush, and nothing at all came out except coolant. If you look at those passages in there they arent that big, you should be able to find everything. Your scoring looks like normal machine marks.
You got very lucky, especially considering where the T stat is in the coolant path. I've never been that lucky, not even once.
The two bolt inspection plate on the back side(Trans side) of the engine on bank 2 is a good place for pieces to be found, too.
So why hasn't anyone invented a filter or catch or something that can be retrofitted into the coolant lines to collect all the broken pieces when this happens. I don't think this would be a hard thing to do.
I did that at least a decade ago.. The issue with it, is it blocks ALL the coolant flow right out of the pump, and you have no idea that you've had a failure. Why is that bad? Because without a flow of coolant, there's nothing to get the temperature up at the coolant temp sender, which is located OUTSIDE the engine, in the front coolant console.
That means you bake the engine, but you never even know it, because the gauge says the engine is normal temp. Lots of good ideas, end up being bad ideas when the whole cause/ effect is looked at in a big picture.
Ah, I see. And this brings another thing into question the OP says the temp was 2/3 the way up but based on this, that would be meaningless since the water is not flowing as well. Got it.
Exactly.. If the impeller fails, no coolant moves in the system. By the time the temp on the gauge moves, in this instance, the coolant in the block and heads is already hot enough to do damage.
In some engines, where this is critical, I relocate the position of the coolant temp sensor, to a place that makes more sense, for a faster CLT response time.
Obstructions, or pump failure can both stop the flow of coolant, or slow it enough that the response time becomes critical. This is why one must always have situational awareness, if you see steam in the rear view, or feel the engine losing power, and you act, the engine can be saved. Today's gauges are no better than idiot lights, because the factory has removed their effectiveness.
Depending on a light, CEL, or gauge warning is often simply not good enough when you have a scenario like this unfold, where the engine gets way hotter than the gauge said that it did.
I would assume you want the sensor to be as close to the heads as possible to give the shortest response time? With a single sensor, I think it's hard to find a location to detect the temp of BOTH heads. Maybe that's why Porsche chose to put the single sensor at a location where the returning streams from both heads are merged (inside the front console)?
It seems that the impeller may become brittle over time - folks with low-mileage 996's are seeing impellers break as well as higher-mileage cars. I'm going to go with a 5-year replacement cycle for the water pump, and I'm basing that on absolutely nothing beyond my gut.
I'll pay it if it means impeller parts stay out of my engine and I'm not up flushing and hoping i got all of the bits (or my mechanic with the hourly rate!)! I have money to maintain the car, and to me making sure your cooling system is up to snuff is part of it. In a lot of cars, it's very infrequent, this car seems to warrant it. Besides my wife drives it with the kids sometimes and they are not getting stranded because I went on the cheap, and I don't expect her to be peering out of the back window looking for coolant clouds or otherwise. The same goes for her car.
People are diligent about dfiferent things.
I just changed the one in my 04 Boxster as it developed a shaft seal leak and it was the original one. No signs of wear on the impeller at all.
I guess 11 years was a pretty good life
Reasonably easy DIY.
Took longer to deal with flushing and filling the coolant then to actually change the water pump.
Probubly take me an hour or so next time without a flush and fill and on the Boxster, you need to mess around with the top and remove a bunch of stuff top get to the engine
Well, BRAVO people BRAVO!!!
I decided to let myself be overtaken by fear and bit the bullet to do the water pump. It is scheduled for tomorrow.
Water pump, low temp thermostat (another controversial topic), serpentine belt, coolant flush. The whole shebang! Plus some fresh oil and filter. (I almost wrote oil and vinegar; I guess I'm still thinking about lunch.)
Bastards!
I recently completely dismantled and rebuilt my M96. The P.O. had replaced the WP+thermostat. Dismantling showed some superficial scoring of the WP housing .The big surprise was finding bits of his old plastic impeller scattered throughout the engine coolant system.There was quite and accumulation in the oil/coolant heat exchanger. Had the c/s bearing not failed, the cylinder head would have overheated anyway.
So how do you know if you found all the broken impeller pieces ? You try to reconstruct the damaged impeller with the bits you recover. Yup, it is nonsense to think you will find them all.
Metal impeller has it's own issues. So precautionary replacement seems the only intelligent solution ?
So it is consensus that you replace with an OEM design with the composite impeller and figure on routine replacement or a different design with brass or steel (or whatever) impeller? What are the issues with a metal impeller?
I replaced mine as PM when I dropped the engine for other stuff. Mileage was @56,000 and pump looked fine but better safe than sorry. I would drop the engine at no more than 60,000 miles and do all the usual PM stuff.
If a water pump with a metal impeller fails the way mine did, the block will be damaged and require repair.
The bearings on my water pump became so loose that the shaft was running really crooked. That broke up the plastic/composite impeller that I had. With a metal impeller, it would dig into the block and damage the face. This is a big problem because there must be a close tolerance between the impeller and the face of the block in order for the impeller to work properly (to have the proper flow of coolant).
I would expect the longevity of the impeller is determined by how good the coolant is in the engine and if the coolant is flushed, replaced when it should be done.
Doesn't the coolant lubricate the impeller and the cooling system?
Looking at Porsche's maintenance schedule, there is no mention of ever flushing the coolant or replacing the water pump. The only maintenance on the schedule is to check the level and antifreeze protection level.
Yes, but the service manager will try to sell you a coolant flush service within 30000 miles of ownership. Replacing the pump, again, is a feel thing. When engines are readily available and don't cost more than the car itself, I usually replace the water pump when I hit 100000 - 120000 on the clock - good or bad. But with these cars it can be a very costly gamble.
Longevity is directly linked to how much the impeller spends in really hot water I think. Composite will deteriorate over time and even faster in heat.
Every 3 years, despite mileage. The water pump is in service all the time.
No, the mfr doesn't always tell you what's best... If they did it would hurt their ratings. The more it costs to maintain the car correctly, the more it costs to own.
It's really simple, you should either test(PH strips or voltmeter) your coolant's effectiveness every couple years or replace it. And preventatively replace your WP on a routine basis with nothing but an OEM unit. The composite construction of some aftermarket units are inferior to say the least. The word "routine" being highly debatable here.
And thanks, Jake, for taking the time to provide such useful information for everyone... your contributions in this thread, and the many others, are greatly appreciated.
Does anyone know what the "plastic" impeller is made of on our cars?
"Many claim that 'plastic' Audi water pump impellers are junk. Why?
A) Earlier impellers were composed of a plastic called Durethan. In the high temperature environment of a coolant system, it encountered reliability issues. This plastic was not heat resistant enough at temperatures over 212� F (100� C). Since engine operating temperatures function in the 190-197� F range (87-92� C), this pushed Durethan to it's designed heat threshold. It increased the risk of the impeller cracking or slipping on the shaft. This is why so-called 'plastic' water pump impellers have developed a bad reputation.
Thankfully, European car manufacturers, including Audi, don't use Durethan water pump impellers anymore! Starting around 2000, a new advanced material called PPS was progressively integrated into Audi water pump impeller design and now nearly all new Audis and VWs come from the factory with PPS water pump impellers."
http://www.blauparts.com/audi_water_pump_replacement/audi_water_pump_plastic_impeller_controversy.html
Very informative.Thanks for the tech contribution.
SO which of the current Porsche replacement WPs are PPS ?
And how can you tell? Is the impeller marked "PPS" or Polyphenylene Sulfide?
For example I found this on Fleabay:
"VW OE #: 07K 121 011 B - PPS Impeller "or "OE #'s: 06A121012E / 06A121012G / 06A121031C - PPS"
So it is easy to find a PPS impeller for a VW/Audi .So what about an M96?
This is the reason why I advocate to use ONLY OEM water pumps. With aftermarket, you may not be getting this updated material in the blade construction.
Rubik nailed it.
Show me an ad for any M96 WP with "PPS" mentioned.
I previously listed all the VW/Audi WPs that are PPS. None from Porsche.
Euro recycling laws require the plastic parts be marked with the type of plastic so "PPS" is a valid identifier.
Conclusion- until the impellers are marked PPS ,keep changing the WP every 50k ??
That's your call/judgement and that's open to a whole another debate which has been going on for ages. I'll leave it at that. For me, I buy from Porsche on something like this. It doesn't make sense to me to save $20-$30 for something I'll probably only change out once or twice at most in my ownership.
WTF is going on - the material they use might break off and block your engine and cause it to overheat - the very thing it's supposed to not be doing.
What did they use years ago on cars?
Because then, usually, all that went wrong was the bearings in them - which you could hear with the screeching noise they made.
Is it because these things are being made cheaper and cheaper and that they do not last as long?
I would think you are better off with metal which can stand the temperature changes better.
Well, I agree with Jake that Water Pumps are in service 24/7, so I prefer a time interval vs. mileage. Every 5 years I'd say. Jake says 3 but that's overkill IMO.
If it's EXACTLY the same part, I have no problems with it if I know for sure. Otherwise if I have to guess at it for something this critical, I'd rather error on the side of caution. I know just because a part manufacturer makes a certain part for a car manufacturer doesn't always mean they are indeed the same exact parts. Sometimes but not all the time.
Well, if it has the Porsche part number stamped in it just like the factory original one that came off my car...
If it looks like a duck
walks like a duck
smells like a duck
pretty good chance it is a duck
I guess you would only buy the ignition switch from Porsche even though both the Audi and VW part are exactly the same. Which would you rather pay, $125, $70 or $30?
Jake,
Explain to me why Porsche water pumps are so junky? When I had my Acura at the shop for the 2nd TB change 6 months ago, I spoke to their mechanics about what fails or what doesn't if you don't do the TB service. Mechanics told me that they almost never see a WP failures, and they have seen cars that went 200k miles on the original WP and original TB with zero issues.
I like genuine parts. Less variables to worry about for DIYers. A genuine water pump is $260 from Sunset. It's a little more expensive but considering how critical the part is and how infrequent I need to change it... Targeting replacement every 5yrs. My 10yr old pump was fine when I removed it but my coolant was very fresh due to a couple of coolant changes thanks to the crappy expansion tanks. I also like to do a complete coolant flush every 4yrs.
German water pumps are ALL crappy (BMW, Porsche, Audi, etc.) The newer electrical WP is even worse...their electronic modules fail.
Acura's WPs use metal blades and they seldom fail like you said.
Compared to Jap, German are not good at making:
- window regulators
- door lock actuators
- water pump
- radiators
- any electronics
The problem isn't impeller material as much as water pump bearings. When the bearings fail, the impellers contact the block, and shatter. The clearance between the impeller and block is often less than 2mm, so little is required in the way of bearing failure to create this scenario.
Yes, we developed a water pump that had twice the bearing strength, but it cost 725.00, and no one would buy it. We made a run and used them on our engines, until we ran out, but never took it any further. Distributors would not consider carrying it, as they could make no money off of it, and buyers didn't think it was even necessary..
Given the look, superb steering feel, handling, unique rear weight bias, driving dynamics, X74, practicality, etc., I'm willing to overlook those pesky parts and any design flaws
Well, my indy just dropped off my car at the office and it looks like my water pump paranoia was in fact warranted. Although, the bearing and blades are perfectly fine, there was a sign of pressure leak which allegedly makes a water pump failure imminent. Also did a full coolant flush and replaced with Porsche pink coolant. Now I also have Low Temp Thermostat. Happy camper here.
BTW: No "PPS" indication on the old genuine or the new Pierburg water pump anywhere. And they are exactly the same.
Rubik: Did you replace just the thermostat or the thermostat and housing? I was looking it up and it appears in many cases it's cheaper to purchase the thermostat and housing vs. just the thermostat. Can't make heads or tails of it.
The problem is a flawed design technically as the belt tensioners are maintaining a constant tension and pulling the water pump to one side placing side loads. This isn't so bad in driving an alternator for example where there can be looser tolerances but when dealing with liquid sealing and tighter tolerances, it can become an issue as it wears.
If you look at some Chevy blocks, their wp is directly driven by the crank. This prevents any side loads as there is no tension on the shaft. Due to space, packaging, and other constraints, this isn't always feasible in all engines though.
OMG, I'm contributing to the typical analysis paralysis and bench re-engineering of the RL mentality here...
He also replaced the serpentine belt and oil change but I gave him the oil and filter. $900 excluding oil and filter.
He also cleaned my air mass sensor and throttle body, took care of a vacuum leak, and checked the suspension. The car is in ship shape.
Amazing information and special thanks to Jake Raby for his input. I live in Japan and am searching for the right 996. The 914s here are wayyyyy out of any sane person's price range
Anyway, I have learned so much from all of your experiences and it has made me a more informed buyer. I have several candidates and hope to pull the trigger soon. If I can step on the soapbox for a moment, please support the shops and businesses that support this site! Their time and advice is graciously donated and the few dollars more you might spend keeps them in business and our cars on the roads.
And the temptations begin! I'm sure it's fitted with a MassIVe power plant? If only the import and registration laws were more favorable. Brought my 1957 MGA here and lets just say you have never seen so much red tape
Regarding your search for a 996, if you're looking for a 99, look for a build date prior to 4/99. I was told that's when the engine problems began (that's when they began including engines that were resleeved or otherwise repaired in order to keep up with demand).
Mine is a 3/99 build, so I'm probably in good shape. We'll find out for sure when they (Flat6) pull it apart to clean it out.
Exactly what I'm doing.
More details to come in this thread, but it may be a week or two until I'm able to ship it off to them.
I don't know if that's an old wives tale about the production date or not....but I had to go look at the build date on my car and it's 2/99.
I will sleep a bit better tonight, but probably only for one night.
Well, I'm convinced. I just ordered the Pierburg pump and Porsche gasket from Pelican Parts for $252 shipped. Now to find good DIY's for replacement and coolant change.
Having owned and maintained our 1990 VW Westfalia Camper with power door locks and windows for 18 years (219Kmiles) you hit the bulls-eye-
Compared to Jap, German are not good at making:
- window regulators
- door lock actuators
- water pump
- radiators
- any electronics
+ goes through half shafts like they were wiper blades.
The
Pierburg water pump
is $210.96 and the
gasket
is $3.79=$214.75 and that would include free shipping. That's without the 10% off that I offer on here, so after that discount it would be $193.28 shipped to your door within the US.
-Jason
straight from the horses mouth...
might be no tax as well
I shopped around when I bought my WP/Thermostat/coolant a few weeks ago and in addition to the best price, Jason is incredible in customer service and I had my parts the next day
Damn. I Just checked and my order from Pelican has shipped already.
Just checked the Warehouse33 site for the low temperature thermostat and the shipping fee for a $46 part is $20. Unless I'm reading that wrong.
I'm here, I assume you're looking at part # 996 106 013 59 for $42.15? I did a shipping quote to my zip of 64111 and I'm seeing $11.17 for ground & $15.68 for USPS Priority Mail.
What's your zip?
-Jason
It said $19 and change for UPS ground in the cart. No worries, Pelican added the part to the order even though the pump was shipped out already, so free shipping.
I will check out Warehouse33 in the future. Thanks for the information. Live and learn.
