NHTSA Update on CR HPFP failure investigation

HPFP VW Jetta 62,200

Hi there,

This might not be the right thread. Please forgive me if not
Direct me to the right thread please.

Not a happy one here. Looks like my Jetta falls into that .11 percent of those VW who almost killed the driver. My wife, lost power right in the middle of a very busy intersection. The car died couldn't start it .

Got the code off it and it points right at the fuel pump. It was late on Friday and I was out of town. Called my bro and he towed it to his shop. There it sits until I find out how VW is going to react.

Any suggestions? Also, if there is a class action on this please let me know.
If you need to join one write me atmrat@g(eeeeee)mail.com


Thanks for you're time

Tom

cashtdi said:

Any suggestions? Also, if there is a class action on this please let me know.
Tom

Click to expand...

I do not know of any class action in place, because so far VW has been paying 100% for fixing all HPFP failures even on cars out of warranty (usually a complete fuel system replacement).
Suggestions:

1) Make sure it was an HPFP failure. There are many possible causes for the car dying in the middle of an intersection. One simple check you (or your dealer) can do is to remove the fuel metering valve on the top of the HPFP (2 torx screws) and look at the screen on the bottom. If is full of little pieces of metal, your HPFP has self-destructed.

2) IF your HPFP does turn out to be the cause of your failure, do 2 things:
a) There is a thread here which is the official place to record all HPFP failures. Enter your info here: http://forums.tdiclub.com/showthread.php?t=299854
b) Go to the NHTSA webpage and submit your case. There is an ongoing investigation and the more complaints they get, the more likely they are to do something. Be sure to emphasize that this occurred in the middle of a busy intersection and that it was a dangerous situation.

Have Fun!

Don

P.S. This was probably not the best place to ask this question but HPFP info is already spread all over the place here, so I'm not certain where the best place would have been (maybe the Mk5 forum?).

Complaining here isn't as effective as ...

> Looks like my Jetta falls into that .11 percent of those VW who almost killed the driver. My wife, lost power right in the middle of a very busy intersection. The car died couldn't start it .

File a complaint.
https://www-odi.nhtsa.dot.gov/VehicleComplaint/index.xhtml
Investigation EA11003 is still "open", but NHTSA doesn't typically take action unless there's a material safety issue. So far, the HPFP scandal has just been a catastrophic financial issue for those who don't get covered by VW.

> Also, if there is a class action on this please let me know.

There was a solicitation some years back. I don't think it went anywhere. You could search on it as easily as we could.

bluey said:

Easiest way to fix lubricity appears to be to add 2% biodiesel. Surely would be a simple "green" recommendation and fix for VWoA to line up a fuel company to produce/distribute this and recommend it for all TDIs.

Click to expand...

Errr, wrong answer. Many are using additives and they still fail. Its a ****e setup and dangerous.

Does anyone recall if the NHTSA info provided by VW has a break down by month? It seems that we regularly have a rash of failures around Labor Day, this could just be my imagination but I am wondering if anyone else has "noticed" this.

turbovan+tdi said:

Errr, wrong answer. Many are using additives and they still fail. Its a ****e setup and dangerous.

Click to expand...

I read through the thread and I believe there were only one or maybe two of those that failed running additives.

turbovan+tdi said:

Errr, wrong answer. Many are using additives and they still fail. Its a ****e setup and dangerous.

Click to expand...

psrumors said:

I read through the thread and I believe there were only one or maybe two of those that failed running additives.

Click to expand...

Yeah, I don't recall any that were religious additive OR live in states that have the 2% mandate. It only takes one *extra* low lubricity event to grenade from everything I have read.

waltzconmigo said:

Does anyone recall if the NHTSA info provided by VW has a break down by month? It seems that we regularly have a rash of failures around Labor Day, this could just be my imagination but I am wondering if anyone else has "noticed" this.

Click to expand...

From what I remember, the NHTSA numbers haven't been updated for a few years. Hard to draw any conclusions without the full story.

kjclow---I agree, IIRC, the "investigation" was only for '09-'10 models. I was just wondering if anyone had the made the same "observation" I had. Unfortunately I lost my hard drive back a few months ago and have not figured out how everything works with my new setup. I have no idea why i have this feeling that mid-August to mid-September would show more failures, but I think it would. Remember, this is just a feeling I have and it could be completely INCORRECT.

Mine died on labor day weekend, about 10 days ago. I can think of two 'possible' reasons for that.

1) One of the busiest travel times of the year, in the US anyway.
2) Summer heat perhaps.

Mine died on the freeway (I-15) as I was headed to Las Vegas from So Cal. It was 109 degrees out that day (35k miles, 2012 Jetta TDI). I had cruise control on while climbing a significant uphill grade. So it's easy to believe that I was stressing the system fairly well at that moment.

I filed a complaint with the NHTSA yesterday.

Summer heat? This has been thrown around before but there is no proof one way or the other unless we have official numbers.

This is the distribution of build dates for the reported incidents in the last updated spreadsheet that VW submitted to the NHTSA. The data peter out at the end of 2011, but that doesn't mean there were no more cases.

The peak for early 2011 dates is not what I expected from the posts here.

For quite a few of the incidents the VW and Bosch findings say 'NTF' (no trouble found) and the pump is OK. So there seem to be cars in the table that stalled or threw check engine lights for other reasons, or that were reported as misfueled, but no damage to the pump resulted.

Does this graph represent reality, because I thought there were more cases than those established ?

martiniwg said:

Does this graph represent reality, because I thought there were more cases than those established ?

Click to expand...

No, it represents the data that VW had to work with thru 7/2012.

And here are the failure dates.
I removed the 121 cases where Bosch states the pump is ok, and VW states it is functional and no trouble was found. That left 180 cases. 131 of them have the defect location given as 'roller surface' with the defect type 'ground up'.

What I really don't understand is the reason why VW labels so many of them 'fuel caused or contributed to pump failure', when either no fuel analysis was done, or none of the values reported (scar diameter, viscosity, biodiesel content, and flash point) point at the presence of gasoline in the fuel sample. There is fuel analysis data reported for 62 of the 180 cases, and only 17 of those have flash point below 52ºC or viscosity below 1.9 cSt.
A few of the others have a few percent biodiesel (only four samples have more than 5%, of those one sample has 17%; all four of those have the roller surface ground up diagnosis, so the biodiesel found in the fuel at the time of failure is unlikely to be the culprit). There is one case of an algae clogged metering valve, which could also point at bad biodiesel.

diesl---what is the multiplier of these graphs? the total number of failures was something like 5900 iirc. it looks that VW may have actually received a batch of bad pumps, used in 7/11-10-11 cars. your graph seems to indicate that those have over twice (four times) the standard rate of failure. I hope I am reading this wrong. thanks for putting this together.

Waltz, I'm just plotting the dates from a spreadsheet in this NHTSA case: http://www-odi.nhtsa.dot.gov/cars/problems/defect/results.cfm?action_number=ea11-003&SearchType=QuickSearch&summary=true
Specifically, from
http://www-odi.nhtsa.dot.gov/acms/cs/jaxrs/download/doc/UCM430247/INRD-EA11003-54339P.xlsx
What I should do is remake the plot of build dates after removing the false alarms. It's clear that after VW instructed dealers to send pumps in if there is any hint of possible HPFP failure, they got a lot of perfectly good pumps...
I also do not know what the sales numbers look like; one of the NHTSA requested documents might have those numbers though. That would be the normalization you are asking about I guess.

I must admit I'm somewhat motivated but not super eager to go through this information. I'm just surprised that nobody else has done it yet, and posted it here.

diesl---for some stupid reason (ME) I can not open any of the linked pdf's for the NHTSA site. I lost my hard drive a few months back and am working with a different system. I an going to see someone saturday about my old one and will see if he can explain what I am doing wrong. When you say, "I'm just plotting", does this mean you are still in the process and will update later or that this is just what you did? Is the vast difference I mentioned from the 07/11 to 10/11 real? up to four times the standard failure rate?

sorry i can not look this stuff up myself at the moment.

So I just realized the 180 failures still have some 'no trouble found' cases in it. 142 have the roller surface damage, 28 have some corrosion but are functional, and 10 have valve or other damage that renders them nonfunctional. But I left the selection like that (i.e. the 180 of the 300 that are not perfectly fine) for the moment.
Here are the pump build dates for those vehicles:


And here are the vehicle build dates:

There is still that peak that Waltz mentions. Until now I thought the failure rate went down after model year 2010; that could still be true if sales figures went way up (rate = number of failures/number of vehicles sold).

I'm posting this as I am finding my way through the spreadsheet; this is really not how one should do this, but better than nothing.

Waltz, the data I'm looking at are not in a pdf, it is a real microsoft excel spreadsheet (.xlsx format, INRD-EA11003-54336P.xlsx) that VW submitted to the NHTSA, and that the NHTSA then posted. You should be able to click on the link in my previous post and save the spreadsheet to your computer.

What one/somebody should do is double-check the meanings of the posted data against the NHTSA questions to VW. As I said, the normalization (number of sold TDIs) might be somewhere in the posted documents.

got it now, i have not installed any of the microsoft stuff to my new hard drive yet. I will look for those disks tomorrow. (Dumbass)

added: where is this 180 number coming from?

according to the list, there were 203 failures reported by 9/12 just on this site

http://forums.tdiclub.com/showthread.php?t=293629&highlight=state

Waltz, the 180 is my count of the cases reported by VW in this spreadsheet that are not perfectly fine.
"I removed the 121 cases where Bosch states the pump is ok, and VW states it is functional and no trouble was found. That left 180 cases. 131 of them have the defect location given as 'roller surface' with the defect type 'ground up'."
I'm now thinking that another 28 of those 180 should be removed, since the pump has some corrosion but is functional.

In any case, what is most needed is to convert the number of cases into a rate, by dividing the number of failures by the number of TDI sold for the same build date period.

sorry, without being able to open these spread sheets I am having a hard time understanding. Are you saying that VW/Bosch contend that more than 4500 of the 5000+ failures were not actually failures at all and that the pumps were fine?

Diesl---I just went back and skimmed the thread for the document dump date (12/05/12) discussion starts at post 4453 of this thread. This seems to be the same time you joined, were you aware of this? Anyway, post 4497 cites over 5800 failures, off which +5200 are for unique VIN's or about 600 were 2nd/3rd/4th failures.

added: the vast difference between the 180 number and 5XXX number is why I was asking about the multiplier.

From document http://www-odi.nhtsa.dot.gov/acms/cs/jaxrs/download/doc/UCM430237/INRL-EA11003-54326P.pdf, VW sold about 40/60/60/70k TDIs for model years 2009 to 2012. They received about 580 consumer complaints about 530 vehicles that are HPFP related, and 5900 field reports about 5200 vehicles until fall of 2012.
I haven't found yet how the 300 cars in the historical spreadsheet http://www-odi.nhtsa.dot.gov/acms/cs/jaxrs/download/doc/UCM430247/INRD-EA11003-54339P.xlsx were picked.

thanks for the pdf, that opened right away. I am not going to be able to read the document tonight but on page 3 it seems you are possibly using the number of complaints (A) vs the number of field reports (B), I will have to look at all of this tomorrow and see if I can make more sense of the disparity. Thanks again.

added:if you have time tomorrow take a look at the posts following the document dump (referenced above), you will be interested in the analysis and discussion. I found posts 4694 and 4696 particularly humorous after reading from 4453.

I bought my TDI in November of 2012, so that would make sense. I just went way back in the thread, and there are discussions of the numbers. Maybe I'll read a bit of that first before duplicating stuff that was already discussed two years ago.

As I said, I do not know how the 300 analyzed cases relate to the 5300/5900 field reports.

I found this post on difference between requested and actual fuel pressure interesting; but I'm not sure I want to keep both numbers up on the scangauge all the time.

waltzconmigo said:

according to the list, there were 203 failures reported by 9/12 just on this site

http://forums.tdiclub.com/showthread.php?t=293629&highlight=state

Click to expand...

I just tallied them up the other day from this thread and came up with 260+

thanks tditom.

hasta manana guys