E46 Fanatics Forum banner

Save yourself up to $200 and some headaches - Electric Fuel Pump PM

574K views 2K replies 369 participants last post by  OURMANDAN 
#1 · (Edited)
Fuel Pump Failure Epidemic 2012

So I have been around this forum and many of the other BMW forums and what seems to be a reoccurring theme is after about 7 years is most of the E38, E39, E46 and other similar vintage BMW's not newer than 2006 are have failing fuel pumps at a very high rate. Probably a rate of about 100%, as I say its not if, but when! 5 years later in 2017 fuel pump failures are still happening and probably at a higher rate due to the age on run hours as the E46 is not getting very old and almost to the point of beyond maintenance and actually at the point of restoration.

So a word to the VERY wise, if you have a spare $150 or so and a couple of hours, order yourself a replacement fuel pump, new fuel pump relay and fuel filter and replace them as a PM item.

Here is one good example of someone that mentioned his "Micro sample of 4 E46's" all seemed to have fuel pump failures between 70-140k miles. Note that mileage is really not a complete indicator for fuel pump lifespan, run hours are what matters, but this is hard to capture. Also keep in mind, in general terms the fuel pumps just plain wear out. A fuel pump that has 70-100k miles is likely over half way worn out. When will it fail, this is anyone's guess. Usually when it is very cold or you least expect it to fail! This thread also has some good pictures: http://forums.pelicanparts.com/bmw-3-series-e46-1999-2005/712384-e46-fuel-pump-failure.html

There are really 2 typical fuel pump related failures, but there can also be some more minor pump related problems where the fuel system will drop too much static pressure when the pump is not running. The static pressure drop can usually be caused by one or more of the following: Leaking fuel pump output check valve, Leaking fuel pressure regulator, Leaking fuel injectors (rather rare), or a basic fuel leak due to hard injector O-rings, a loose or damaged flexible hose connection or possibly even a leaking valve a the fuel pressure test point.

One case is where the fuel pump hard fails and does not provide any fuel to the engine at all. This is a pretty cut & dry situation where the problem is rather obvious. A quick and inexpensive way to determine if you are have fuel supply problems is to purchase a $3 can of starting fluid, remove the air filter and give a good, solid, 2 second shot of starting fluid into the intake path/MAF inlet. Then try to start the engine, if the engine starts then almost immediately dies, then you could repeat the starting fluid process, but more than likely the starting fluid test has confirmed you have fuel supply problem. This will most likely be due to a hard failed fuel pump, but it is worth checking the fuses and possibly switch the horn and fuel pump relay just to rule these out as causes of your no start problem. In a bind you can often get the fuel pump restarted with a number of good smacks on the bottom of the fuel tank on the passenger side. If this does not work, you could remove the rear seat bottom and the fuel pump access cover on the right side of the car and use the butt of a hammer to whack on the top of the fuel sending unit. Be careful not to hit the connector, wiring or the output connection where the fuel line is connected. If you are LUCKY, you could get the pump to start again and maybe make it to your destination.

Another failure mode is a soft fail fuel pump that fails to put out enough pressure and/or fuel volume. These are far harder to identify and symptoms may appear to be intermittent or very minor. In some instances the problems may not even seem fuel related. Surging, poor acceleration, possible lean codes, extended cranking when starting, running out of fuel when the gas gauge shows 1/4 of a tank. Soft fail fuel pumps are not really worth diagnosing in my opinion. For the price of the pump and how easy it is to install, I would just plan on replacing your pump if it is more than 7 years old and you have unidentified driveability problems.

In addition to the pump being a problem, fuel filters are very often neglected and over looked and can cause similar issues to a soft failed fuel pump. Please carefully read the following information on fuel filters below.

SPECIAL NOTE ON FUEL FILTERS - since the early 80's fuel filters have much finer filter elements and will catch far smaller particles to protect the fuel injectors. Although many cars have large fuel filters and manufactures claim the filters can be changed every 50-60k miles or are even considered "Lifetime", based upon field observations, I would HIGHLY suggest replacing your fuel filter every 30-50k miles. If you open up any of these filters after 25k miles, they are often caked with a very fine black film. This black film builds up and starts to restrict fuel volume and pressure and can cause unusual engine performance issues. From years of monitoring Forums and my own observations, it seems that many of fuel filters with the black film seem to be cars from the East Coast or North East corridor of the US. This may be due the age of the storage tanks both for the distribution network and the in ground tanks at the local gas station. While this thread is primarily focused on fuel pumps and fuel pump failures, PLEASE understand that replacing the fuel filter is an assumed maintenance item. I do not recall seeing specified service interval from BMW for these fuel filters, they may consider them "Lifetime", but I would HIGHLY suggest you consider changing them every 30-50k miles to be on the safe side.

While checking and replacing the fuel filter, also carefully check and replace the fuel pressure regulator hoses and the "F" fitting in the upper intake boot as any vacuum leaks in these hoses and fittings can also contribute to Lean conditions possibly leading to the CEL lighting. Please note that this "hose" that is connected from the F connector in the upper intake boot and to the fuel pressure regulator on the majority of these cars is NOT a vacuum hose. This hose typically does not have manifold vacuum for most models, this "hose" is nothing more than a filtered Atmospheric reference for the fuel pressure regulator. For most of these models, if you present manifold vacuum to the fuel pressure regulator, the fuel pressure will be reduced by approximately 10 PSI at idle and will lead to driveability problems and Lean codes.

What seems to be happening a lot lately is folks are not replacing their fuel pumps as a PM item, they are getting stranded, requiring tow jobs and/or dealer or rush fuel pumps to be purchased of up to $400+. Some even have to pay for the install as they do not have the tools on hand or the time.

These pumps are very easy to replace, only need a few basic tools and you do not even need to get under the car. Even if you buy the pump and do not get it installed, keep it in the trunk with the tools to change it and a copy of a DIY until you get around to actually installing it.

It is almost as easy as changing a tire!

So these fuel pumps are kind of like death, taxes and the cooling system.

As the Nike commercial says, "Just Do It!"

Your wife, kids, girl, buddies, employer, wallet and your schedule will all appreciate a PM fuel pump job.

www.bmaparts.com seems to have the best pricing on the fuel pump, not sure about the relay and filter?

I think there is also an online discount code that may still be good, something like SpecE30?? If I find it, I will edit this post with the discount code, but I do not guarantee it will work.

UPDATE: The BMA online discount code is SpecE30. Reports are the replacement pump is under $108 with free shipping. NOTE: As of Aug 2012 it appears the pump prices may have increased slightly to closer to $120?? As we are 5 years after this thread was originally written, the Simens/VDO fuel pump and sender assembly has gone up in price, as of mid 2017 the typical going price is around $145.

To help out people that are in fact in a bind and stranded, many of the local auto parts stores are reboxing and offering the Siemens/VDO pumps but they tend to be more expense that you could proactively purchase the part online when you are not stranded. Any example is this thread where it was found that the part was reboxed for a local suppler - https://forum.e46fanatics.com/showthread.php?t=1130362

Members here have had decent luck with Bosch branded fuel pump and sending unit assemblies.

A note about purchasing very low cost fuel pump sending unit assemblies, recently a number of members have purchased from Amazon or ebay pump and sending unit assemblies for as low as $50 and installed these as PM replacements for the original working fuel pump and have quickly found themselves stranded or with driveability problems. Just use your head, how do you justify a fuel pump sending unit assembly that is the cost of the typical fuel filter for these cars that has effectively no moving parts, not electrical components and no wiring. Think about it before you decide to be penny wise and pound foolish and trying to spend less that you would expect for a quality part. I DO NOT suggest you even consider spending less that $140 for a name brand part, this still does not rule out that you could still end up with a counterfeit item. Make sure you are buying parts form a reliable source and supplier. When something is too good to be true there is usually a reason.

Note about soft failures on the fuel pump

Some of the fuel pumps are soft failing. This means they work and you do not have issues starting your car, but they can cause intermittent hesitation, misfiring, stalling as well as problems running out of fuel when the tank is about 1/4 full. This is due to low pressure and/or volume causing the siphon or suction jet pump that transfers fuel from one side of the tank to the pump side to not transfer fuel correctly. What happens is you still have fuel in the non electric pump side, but the electric pump side runs out of fuel. This usually happens with 1/4 of fuel showing on the gauge. The hesitation, misfiring and stalling is usually again to low pressure and/or volume from a weak fuel pump. Many have recently replaced their fuel pump & filter as preventative and found that they have crisper throttle response, better overall acceleration and sometimes a small inprovement in fuel economy, which was not noticed until the pump and filter were replaced. So there may in fact be some hidden benefits to preventatively replacing your fuel pump and filter.

Most of the soft failures are due to severely worn brushes and commutator and/or wear in the actual pumping section of the fuel pump. Think about this analogy. Fuel pump brushes and the commutator are like brake pads and the rotor. The difference is the pump brushes are under constant spring tension on the commutator, it is like you driving the car with the brakes partially applied. Brush and commutator wear is part of the normal operation of a typical DC brush fuel pump, therefore the fuel pump is consumable, similar to a starter motor, but the starter motor is not running the entire time the engine is running.

There is a lot of mis-information from fuel pump suppliers and multiple Internet sources about running the fuel level low in the fuel tank, this information is pretty much garbage. Same hold true for fuel contamination. When you are talking about any fuel pump that is failing beyond the warranty and/or after 8-10 years, these are all failing due to them being worn out. Pump run hours is the main factor in pump lifespan, not vehicle mileage.

Contaminated fuel could damage a fuel pump, most contamination is due to water. But for the most part if the fuel is significantly the engine will run very poorly or not run at all. Once the contaminated fuel is removed and/or diluted with fresh fuel, there is little chance this will cause an imminent fuel pump failure. But if the fuel pump is over 8-10 years old, it would be wise to replace the pump at that point in time.

Some people prefer to wait and replace the fuel pump when if fails, the problem is unlike a dead battery you cannot always get the car started with little time and effort and get the car to a place where you can get it repaired quickly and easily. You cannot always find a fuel pump 7 days a weeks and at most local corner stores. So a fuel pump failure can and will cost you a lot more time and money than say a weak or dead battery.


Fuel Pressure & Volume Testing

From the Bentley manual, following are fuel pressure and volume specs for anyone that happens to have a fuel pressure gauge handy and wants to test either the fuel pressure and/or volume. I have not tried this trick, but some have stated that you can use a regular tire pressure gauge to measure fuel pressure in a pinch. BE CAREFUL DUE TO FUEL SPRAY AND FIRE/BURN HAZZARD!!!

Note on most E46 except the M3 the fuel pressure tap or test point (Schraeder valve) is located on the front of the fuel rail under the upper fuel rail/engine cover. This upper fuel rail/engine cover will need to be removed to access this fuel pressure tap or test point.

On the E46 M3, the fuel pressure tap or test point (Schraeder valve) is located on the fuel pressure regulator block, under the drivers side floor board near the fuel filter.

On SULEV vehicles, I have not done a thorough investigation, however, expect the unexpected with these cars. These cars have the stainless steel fuel tank with the sealed fuel pump. This pump actually has 15 year/150k mile warranty and cannot be serviced without replacing the entire fuel tank assembly which I have heard is in excess of $6000!!! Note to self, if you have a SULEV car sell it before 14 years / 140k miles.

Fuel Pressure Specs Per Bentley
All engines except the S54 3.2L M3 - Fuel pressure 3.5 +/- 0.2 bar or 50.76 +/- 2.9 PSI
Residual fuel pressure after 20 minutes - Fuel pressure >3 bar or >43.51 PSI

Additional note/clairfication for the M52 (323/328 models). These models may utilize a manifold vacuum controlled fuel pressure regulator unlike the M54 (320/325/330 models). Bentley did not properly or clearly outline how to test the fuel pressure on the M52 (323/328 models) from my perspective. For the M52 (323/328 models) the 3.5 +/- 0.2 bar or 50.76 +/- 2.9 PSI fuel pressure is the stated spec, however, I would suggest checking the pressure with the vacuum line to the Fuel Pressure Regulator BOTH connected and disconnected. If the fuel pressure is tested with the engine running AND the fuel pressure regulator vacuum hose connected, the fuel pressure could be approximately 2.8 bar or 40 PSI IF the fuel pressure is actually vacuum controlled. Please make note of this clarification and report what behavior you see with the 323/328 models.

Unfortunately I do not have ready access for many 323/328's to verify if there is any vacuum present at the FPR. The 323 FPR hose is connected to the main body side port of the CCV and I know in a number of cases if there is vacuum present at the CCV main body side port. Not sure if this is with a good CCV or a failing CCV, but any sort of vacuum presented to a FPR will lower the fuel pressure even if it is a FPR that is designed just to have atmospheric reference like the 325/330 models. Might as well check this while you are under the hood checking the fuel pressure on the 323/328.

S54 3.2L M3 - Fuel pressure 5.0 +/- 0.2 bar or 72.5 +/- 2.9 PSI
Residual fuel pressure after 20 minutes - Fuel pressure >4.5 bar or >62.2 PSI

Fuel Volume Specs
All engines 1.12 liter (1.16 qt) for 30 seconds at 12 Volts or while crank


If your car dies on you with 1/4 of fuel, you can sometimes get it started again by adding a few gallons of fuel. But beware you will again likely run out again once the gauge shows 1/4 tank. This may get you going if you are stranded and buy you a bit of time so you can shop your parts and DIY the pump. If you access the Hidden OBC menu (Hidden OBC Menu - Check Voltage, Temp, Fuel Level - https://forum.e46fanatics.com/showthread.php?t=239619) you will be able to check the fuel levels in each side of the fuel tank. Note that most all BMW's have a kidney shaped fuel tank with 2 separate fuel level sensors and 2 fuel pumps, siphon/passive pump the drivers side (Left side), which rarely fails, and the primary electric pump on the the passenger side (Right side).

The way these BMW kidney shaped fuel tanks function with the 2 fuel pumps is as follows:

The electric fuel pump on the right side of the tank is constantly running and excess fuel is returned from the fuel filter/regulator back to the left side of the fuel tank. When the fuel enters the left side of the fuel tank if passes through the passive siphon pump and transfers fuel from the left side to the right side of the fuel tank. Once the fuel level drops below about 1/2 tank the left side of the tank quickly drains and all the fuel is kept in the right side of the tank. If you access the Hidden OBC menu you will be able to see the fuel transfer behavior and if you see the fuel level in the left side of the tank higher than the right side of the tank once stabilized, it is likely indicating a soft fail fuel pump. But also note there are also fuel level sending unit failures that can also show similar behaviors. Best to get to know your cars total average range using the trip odometer and then use this along with fuel level to determine when you may have to refuel and/or have a potential problem cropping up.

It is feasible that a bad, intermittent or dirty fuel level sender could lead to incorrect fuel level readings and the car to run out of fuel while the gauge still indicates fuel is in the tank, this is a rather rare situation. If you are tracking your distance driven per tank, you should realize quickly you are exceeding the typical distance per tank you would normally drive. From my experience to date, I have not seen many sending unit failures and usually the sending unit failures or the few instances a sending unit float gets fuel saturated, the sending unit usually under reports the fuel level rather than over reports the fuel level. Here is a good link with pictures of both sending units removed from the tank.

https://forum.e46fanatics.com/showthread.php?t=738747

As for actual fuel pump failures, there are many myths and ideas as to why and how the pumps fail. Also suggestions that the pumps are overheating due to running the tank low on fuel for extended times and possibly the pump drawing air from time to time. But I can say most of these suggestions are mostly myths. The fuel pump is a DC brush motor that is constantly running and circulating cool fuel. The failures of these pumps really have nothing to do with overheating but simple brush and commutator wear just like a starter motor for an engine. What is typically happening is the brush fail to make proper connection with commutator and fail to start. Most of the fuel pump failures tend to happen after the car is shut down, there appears to be a very low percentage of failures while driving.


Also a few helpful hints on fuel pump replacement -

1. Wiring Connector
Many people new to BMW's or German cars may not understand how to release the wiring from the fuel senders. The connectors have a slide release that is a bit tricky but will be clear once you study it for a moment.

See attached pictures of the wiring connector up close both "latched" and "unlatched"

Auto part Technology Wire
Tire Auto part Automotive tire Wheel Automotive wheel system
Wheel Auto part Tire Automotive tire Automotive wheel system


Pictures courtesy of Frank_E46


2. Make sure you install the gasket on the fuel tank opening lip and not the sending unit
. When you remove the sending unit, the round gasket may come out stuck on the sending unit or you may partially pull the gasket out of the tank when removing the sending unit/fuel pump assembly, however, you need to install the gasket on the fuel tank lip before reinstalling the sending unit/fuel pump assembly, there is a small grove in the side of the gasket that goes around the inside lip of the fuel tank opening. What I actually do is install the sender with the gasket loose on the sender, then once the float is inside the tank, I then put the gasket on the tank lip before seating the sending unit. Also be VERY careful to install the gasket the gasket in the correct orientation when you install it on the edge of the fuel tank. This gasket is asymmetrical and should only be installed in the proper orientation. I believe there is the word "TOP or UP" embossed in the fuel sender gasket. If you install the gasket upside down you will have a fuel leak. Fuel leaks will only likely occur at 3/4 tank or higher. Once below 3/4 tank, you will likely not have any leakage.

If this gasket is not installed properly you may have also have Evap codes triggering as well as anything from a large fuel leak, to just a heavy odor of gasoline.

So be warned, make sure you note the orientation of the fuel sender gasket when you remove the sending unit/fuel pump assembly and reinstall the gasket in the correct orientation and make sure the gasket is fitted inside the tank opening and groove in the side of the gasket is installed on the edge of the fuel tank opening.

3. Fuel level can be an issue. If you are 3/4 tank or below you are in pretty good shape, but if you just filled up and need to change the fuel pump, you will have a problem with fuel spilling out of the tank onto the top of the tank and the driveway or asphalt below. Remember gasoline is not only flammable, it can soften asphalt and be an issue if you happen to be indoors. It is apparently very hard if not impossible to siphon fuel from the tank via the fill pipe, so be forewarned and this may be another good reason to proactively replace your pump on your schedule and not the cars schedule!

4. Be careful about residual fuel pressure in the fuel line before changing the fuel pump or filter. If your pump failed and your car will not start, you likely will not need worry about residual fuel pressure, however, if the car still runs and you plan on changing either the fuel pump or fuel filter, it is wise to remove the fuel pump fuse, start the car and let it idle until it dies. If you plan on removing the fuel pump, you can disconnect the wiring connector to the right side fuel sender and start the engine and let it run until it dies. NOTE, if yo disconnect the fuel sender, you may end up with some trouble codes for the fuel sender and you fuel level and mileage figures may be WAY off until you fill the fuel tank. The trouble codes will not likely show up with a standard ODB scanner, however, INPA, SSS or BMW Scanner may read these errors. Just make a note if you do run the engine with the fuel sender disconnected in case you see any codes, you likely forced them to be triggered. This is way you can pull the fuel filter or pump without a high pressure spray of gasoline. You will still likely get some fuel leakage, but hopefully not under pressure. It might also be wise to remove the fuel filler cap to also release any residual pressure in the fuel tank, but suggest you reinstall the cap before removing the fuel sender/pump combo.

Additionally I have no hard data, but it appears the 2002 and newer cars may have more of an issue with the pump life as compared to the very early E46's? Also case in point, many of the newer turbo BMW 3 series have had a series of fuel pump failures while most of these cars were still under warranty. So just making the comment that the pump design can possibly influence life expectancy.

Do not say you have not been warned!

Denial is not an option, as you will get a rude dose of reality.

Wanted to update this thread with some failure mode information.

These fuel pumps are all dying due to worn brushes and commutators, very little to do with contaminated fuel. Fuel contamination can accelerate fuel pump wear, however, the fuel sock is actually very fine and will trap or block all but the smallest particles from entering the fuel pump. Additionally almost every "fuel pump" at your local gas station has a large inline fuel filter that looks like a large canister oil filter. These filters trap quite a bit of debris that may be in the gas station in ground tanks or distribution lines. The fuel pump is a standard 2 pole/brush DC brush motor and the brushes and commutator wear out, probably faster than they should. This may be partially due to changes in fuel or fuel additives, but who knows for sure.

I took apart the pump that failed in my E46 at approximately 9 years with 84k miles. The brushes were shot and the commutator was HEAVILY worn. One brush was worn all the way to the copper stranded wire that is embedded in the end of the brush, about 1/2 way more worn than the other brush. I am guessing that the brushes were worn down by over 3/4 of their original length. Not sure if the uneven brush wear was due to the way the pump is angled in the sending unit or why this happened. In addition to SEVERE brush wear, the commutator was unbelievably worn, probably over 0.110" of wear or almost 1/8" diameter of the commutator was gone!

All the worn commutator copper and brush carbon is probably part of what ends up in the fuel filter and may be some of the black color that many find when changing the fuel filter.

On to fuel pump longevity.

Overall vehicle mileage or age has VERY little to do with the pump life expectancy. Keep in mind for every second the engine is running, the fuel pump is running as well, or just slightly longer if you consider the fuel pump priming. Pump run hours and the total number of pump rotations is what actually determines when the pump will fail.

It is assumed the pump life is around 5000-6000+ run hours, but this does not seem to match up very well with the failures we are hearing about. You could possibly tie this run time to mileage, bit is VERY hard and not very accurate. I am sure someone may challenge my info here that may be a Product Engineer with system life expectancy calculations. All that being great, but I think there is enough real world data to show that the majority of these fuel pumps are lasting around 4000-5000 run hours. Maybe the design life was longer, who knows. But the failure rate is pretty high over the past few years for E46 fuel pumps.

Estimating fuel pump life expectancy, you typically have 2 primary kinds of vehicle usage. There may be some middle ground, but the middle ground is based on the 2 primary vehicle usage profiles. Highway Warriors that can put 20+k miles on the car a year and the Urban Dweller that might only get 12k miles or less on the car a year seem to make up the main building blocks of vehicle usage.

My E46 is clearly an Urban Dweller. It probably averages about 90 minutes a day driving, possibly a bit less. The fuel pump in my E46 failed around 84k miles as I recall at around year 9. Doing the math, 90 mins x 365 days x 9 years/60 (mins in an hour) = 4927.5 run hours on the fuel pump! Notice I did not factor mileage in the equation at all because mileage has little to no bearing on the run hours. If I was to calculate the Average MPH over the 884k miles, with the data above, this works out to 17 MPH average speed for the life of the vehicle. I think in reality using the Trip Odometer most of the time, the Average MPH was usually around 22 MPH for the type of driving the car sees weekly. 22 MPH Average Speed may actually be yielding more along the lines of closer to 4000 hours of run time, which in the big scheme of things in NOT very reliable and not a very long life expectancy for a fuel pump.

I think most observant people would see a Highway Warrior could clock up far more miles in the same 5k hours if the cars speed was averaging 30-40+ MPH rather than the 20-22 MPH that my E46 would run.

So for instance, a real world situation I deal with weekly is I have a trip during the worst part of the day for traffic. I drive 12 miles and it takes me 1 hour. So this means I am averaging 12 MPH for this trip! Not very good a all and remember for each trip like this, I am adding 1 hour of run time on the fuel pump as well as other parts of the vehicle to go only 12 miles!

For the person that had a garage queen that is 10+ years old and has 50k miles on it, I would expect the pump to have a fair amount of life left because the pump probably has not seen a higher value of run hours.

There is some thought based on discussions that the earlier fuel pumps, pre- 2001 may last longer than the 2001-2006 fuel pumps. Not sure this has been proven with hard data, but this is a theory that is circulating around a bit.

But I think in general the fuel pumps in these cars should be replaced around year 8-9 if you expect to have reliable transportation. Some may disagree, just update the forum when your pump fails and you are left stranded. It is just a matter of time.

Links below to pictures of failed fuel pumps cut open showing the worn brushes and commutator. In both of these cases, the fuel pumps would not run on morning vehicle starts. You can see how thin the one brushes is where the copper braided wire was running on the commutator and also notice how severely worn the commutator is on both fuel pumps. The commutator material is worn over 1/8".

Fuel pump from a 2006 E46 with 84k miles that was used as a city commuter car that would never average more than 20 MPH per tank of fuel cut open to show worn brushes and commutator -

https://www.dropbox.com/sh/tdx5kech2h6mvha/AADAzsixZq5PcF63WsRmk2Pxa?dl=0

Fuel pump from an original owner 2000 E46 with 155k miles that was used for a primary car for the first 10 years of if the cars life, then the last 7 years this was a secondary back up car with very little miles put on it. Point is this car probably encounters close to 14k+ miles per year annually, mostly highway driving/commuting. So expect the fuel pump run hours to be about similar to the fuel pump above.

https://www.dropbox.com/sh/kh9u9n9i8rwklg8/AACwHq0lMDRoDMKKp3wNpu5Za?dl=0


A lower hour/mileage pump that was pulled from a parts car. Probably only a few years on this fuel pump. Note the minimal armature wear when compared to the failed pumps.

https://www.amazon.com/clouddrive/share/UsrpMxsDwZ1Wd4QRrxSjPgvy3QymuKhAnz1Bw9kbjub


Also a very interesting testing case and analysis of fuel pump life and impact of different fuels. Interesting that brush and commutator wear appears to be greater with gasoline than with E10 or E20 blends.

http://www.agweb.com/assets/import/files/e20endurance.pdf


Regarding all of the claims of not running the vehicle below 1/4 tank for fuel pump cooling, this is pretty much a bunch of BS. Most all modern electric fuel pumps have the fuel volume of fuel passing directly through the center of the fuel pump which actually bathes the commutator, brushes and bushings in fuel. So as long as the fuel pump has a supply of fuel, it is being cooled. Additionally I have yet to see a high mileage heat related fuel pump failure. I have cut MANY fuel pumps open for inspection and the majority of the BMW/German design fuel pumps that have failed had SIGNIFICANT commutator and brush wear and it was clear this commutator and brush wear was the root cause of the pump failure.

DC brush motors are CONSUMABLE and they have a designed or finite lifespan. Maybe they should be lasting longer than what we are seeing here in the real world, but like anything, the real world is what we all face on a daily basis. I think the side commutator/brush design pumps may wear faster than the end commutator/brush design fuel pumps.

Should listen closely from the 9 minute mark. I think the Airtex Director of Engineering should be a pretty reliable source. Far more knowledgeable the majority of "expert" mechanics which I have met VERY few of.

What is interesting is if you watch the video and listen to the discussion regarding the pump commutator and the overall issue of wear concerns, there is NO mention of motor brush wear. It may be due to the DESIGN of the end based commutator and the way the brushes ride on the end commutator. I expect that the end commutator/brush design is far more robust when compared to the side side commutator/brush design that the E46 and many other BMW and other German model vehicles use. There may be material differences as well when compared to the BMW/German design.

But for the BMW pumps that I have cut open, they ALL have significant brush and commutator wear and I have not found much problem with the actual pumping portion of the pumps, more often than not the pumps slow down due to commutator and brush wear and also fail to start due to extreme commutator and brush wear just like a worn starter motor that have have personally rebuilt hundreds of over the years.

https://www.youtube.com/watch?v=piOzN1tZI_E

https://www.youtube.com/watch?v=lMVvuME22M0&feature=youtu.be

https://www.youtube.com/watch?v=QbzEeWy1b3o
 
See less See more
3
#425 ·
Are 2006 models outside this issue? I have a low mileage (42k) E46 M3. If the fuel pump is one of the problematic ones, I'd rather replace it now than worry about a failure.

The owner's manual actually states damage can occur if below a range of 35 miles of fuel left which is interesting. Never had a 'low fuel' = damage warning in any of my owner's manuals out of 10+ cars I have owned.
 
#427 ·
I believe the M3 pump is very much like the M5 pump?

The M cars do not have an integral fuel pressure regulator built into the fuel filter. They also run a much higher fuel pressure, I believe up to 70-75 PSI?

The issue with running the tank too low is risk of fuel starvation during hard corning and possibility of leaning out the fuel mixture and causing the combustion temps to quickly climb.

The M cars also have a fuel baffle right at the fuel pump pickup to help reduce fuel starvation during hard corning.

The fuel does a few things to help out the pump and injectors.

The fuel actually lubricates the fuel pump and injectors and also cools the pump and injectors.

If you starve the fuel pump too much, you risk reduced cooling and lubrication as well as putting the engine in a risky situation where the combustion temps can quickly rise and cause knocking/detonation and possible engine damage if continued to be driven hard.

I also believe the M pumps may not last as long as the non M pumps due to their increase pressure and more demanding application??

This is my point of view based more on the E39 M5 as I have had my hands on many more of them than the M3.
 
#430 · (Edited)
He likely has his original DISA, CCV, cooling system and battery too!!

Do not worry the fuel pump lightening strike will get him at -10F when he has to go to the job interview first thing in the morning and there is 10 inches of snow on the ground!

He might be a bit luckier as it does appear earlier pumps may be a bit more robust, but they are not so robust to last forever!

It is a DC brush motor, it is destine to fail due to commutator wear at some point!!

I am sure we will hear the story that it happens at the most convenient time with only 1/2 tank of fuel, the car is in his driveway and BMA parts gets him the parts the next day by 9 am without expedited shipped!
 
#433 ·
Sounds like a likely suspect. $3 can of starting fluid will answer the question for sure. Only the M cars may trigger a code and it is usually for the relay when it is actually the pump!

Read the first post in this thread.
 
#439 · (Edited)
wow this fuel pump issue isnt a small one. im going to replace mine now to. its still original. i wanted to replace it as the fuel level senders is crazier than a drunk guy drinking metholated spirits. but after reading all this i just gotta say thanks that its still working! cars got 220000km on clock. chit i gotta replace it soon hahaha.

i beleive in preventative maintanance on a car. cuz who likes to get stranded and think to yourself, wow those guys were right!!
 
#441 ·
Just to update everyone. After getting my pump and filter in, the car is a lot smoother between gears. When rpm drops for clutching and then accelerating, the whole thing seems a lot smoother- no herky-jerky. I thought I maybe had a VANOS problem as well, but everything seems pretty good now.

I hear the smoothness might come from replacing the filter more than the pump, but my pump was super-loud before it finally died, so I'm sure it was not doing great.

Yay for new pump and filter.
 
#442 ·
Suggest 30k fuel filter interval changes for optimum performance as well.

You should see all the crap that gets trapped in these filters by 60k miles!
 
#444 ·
Might be a regional issue, but everything I have take apart or looked at in my area is caked with the fine black powdery substance.

Likely something coming from the tank farms in our area??

The pumps are also supposed to have spin on oil filter looking post filters, but how often do these get changed and how fine are they.

I think I calculated about 2500-3000 gallons goes through the average filter in 60k miles???
 
#447 ·
I admit I didn't have the patience to read 23 pages of posts! However, one thing to remember is that your in-tank fuel pump is cooled by the fuel, and overheating is an accelerator of pump wear. if you refill your tank when you get down to a quarter tank, you will keep enough fuel in the tank to cool the pump, and minimize the amount of sludge that gets sucked into your pump from the bottom of the tank. 111,00 mile and still on the original pump.
 
#458 ·
Many people mention this every day on here (sorry, you are not the first), so I am not picking on you specifically, but people always give this advice about not running below a 1/4 tank because the pump will supposedly pick up more dirt, but the intake for the pump is in a fixed position at the bottom of the pump. By that logic, it is always sitting in the dirt. I am also not sure what sort of dirt people are expecting to find in the bottom of the tank, but mine was clean as a whistle when I changed my pump. I did not see anything settled on the bottom. The screen was also perfectly clear of debris. It was just discolored, while the new one was bright white. This was after 120k miles. Does anyone have a pic of an e46 tank with a lot of junk at the bottom of it? Can anyone say for sure that they saw a lot of crap in the bottom of their tank?
I can agree with the pump needing to be cooled, so you need to keep some gas in there. But this gas tank sludge is somewhat mythical, and if it does exist, then the intake is sitting in it if you are on empty or completely full.
I know it was a problem a long time ago, due mostly to the materials used to make the old fuel tanks, which would start flaking off or the tank would corrode from the inside out. Also, the old independent gas stations were not the cleanest places either. But with the large conviencence store/gas stations, most of them are relatively new constructions, with relatively new modern tanks to store the gas. I just dont think large particle sized dirt is an issue with our gas any more. Small particle dirt I can believe, and that is what the fuel filter is for. I think this is just something that gets passed around and accepted as common knowledge, and no one really questions it.
Cleaning the wheel speed sensors is another good example of this. They are a magnetic pickup, which is not affected by dirt. That is why they use them way out there near the tires and brake dust and dirt. I doubt there was ever an optical type wheel speed sensor, which is the type that may benefit from a cleaning. If there was, I bet it was not used for long. But people still advise this on here every day.
 
#448 ·
I was wondering if I could maybe get some help.

I replaced the pump and filter, and the car seemed ok.

Now there is a gurgling sound coming from the tank and I'm using a lot of gas. I put the tank up to 3/4, after driving 10 miles I am at half a tank. There are no codes on the cheap e-bay scanner I have and the car is very smooth.

What might be causing the gurgling sound and loss of mileage? If I should start a new thread for this I can do that.

Thanks,
 
#460 · (Edited)
I would think that the malfunction would be with the sending unit under the left rear seat, which transfers fuel between the 2 sides of the tank. It has a "sucking jet pump" on it, which could be the gurgling sound you are hearing. Also, the rapid drop in fuel level also points to this. If it is unable to transfer fuel from the left side of the tank, then you are only using fuel from one side, so what would have been 1/4 tank is now 1/2 of the total fuel available. Since the level is measured on the pump side, this would make sense.
It would be parts #2 and #3 in the diagram in the link below:
http://realoem.com/bmw/showparts.do?model=BN53&mospid=47713&btnr=16_0373&hg=16&fg=15
It is possible something just pulled loose. It is worth checking at least, before buying parts.

Edit: This actually makes a lot of sense. If something has failed over on the right side, it could be cycling continuously, trying to adjust the levels, which would explain the gurgling noise. IIRC, there is something in the hidden menu that allows you to see the level in each side of the tank. I would check that, and see if one side is not higher than the other by a large amount. Check jfoj's link for the list of hidden menu items you can access.
Also, I was not clear about this, but our cars have a split tank, and a pump in between to keep the levels even. I believe your malfunction is with the system that moves the fuel from the left tank to the right one, which keeps everything level. This is because a single tank would be far too simple for BMW to give us. Not enough possible failure points. :)
 
#451 · (Edited)
Hi, thanks for your reply.

No visible gas leaks anywhere. It was parked overnight for 2 nights, then all day at work. No gas spots anywhere below it. I went out just now and touched every connection from the tank to the filter and everything is dry. No gas smells outside or inside the car. I have not yet taken off the panel under the seats to check the pump, but like I said nothing smells.

I haven't hit anything, drove non-energetically over the weekend and to work today for a total of maybe 30 miles.

The gurgling can be heard inside the car. When I put my hand on the tank I can feel it bubbling. I double-checked my connections to make sure I plugged the hoses back in the right way on the filter. The blue hose in the middle hole, the other one in the outside hole.

Everything is clean and dry under the car.

Yeah, I love the color; I had to drive out of state to pick it up. Haven't seen the graugruen thread yet.

Thanks again
 
#450 ·
330ci are you really using a lot of fuel?

Are you leaking fuel, this may be more likely, fuel pump seal will usually leak over 3/4 tank. Easy to pull rear seat again, pop access cover and see if you have a heavy fuel smell.

Could be a fuel sender mis-reading as well.

Suggest you read the first post in the 2nd link in my signature and pay special attention to the fuel pump gasket comments.

The read the link for the Hidden OBC Menu and you will be able to see both fuel level senders and see how much fuel is in each half of the fuel tank.

Then also check your total distance you are able to safely drive. Most E46 can get close to or slightly more than 300 miles per tank.
 
#459 ·
Basically, all sludge that enters your tank is going to exit through the fuel pump, and (hopefully) stop when it gets to the filter. As MJ is saying, that's independent of the fuel level, since the strainer is permanently sitting down there in the sump.
 
#461 ·
The sucking jet pump stuff is on the left. Both sides have level senders, and their outputs are added together to read total fuel level, so the gauge on the dash should read the same whichever side the fuel is in. If they both work. You can read them separately in the hidden menu.
 
#464 ·
Yep. I got my right and left confused. O__o
I corrected it. You are too quick on the replies! :)

Sucking jet pump actually extends from one side of the tank to the other. So it's on both sides.
Well yes, because of the whole balancing the fuel load thing. I was fixing that too, but again, you guys are responding too fast! Give me a minute to read what I wrote, and sort out the brain farts!
 
#463 ·
Been a long time since I replaced an E46 fuel pump, but on my E39 M5, the pump had to be seated correctly as there was a round pipe with an O-ring that came from the left side siphon pump as I recall?

Is there anything like this on the E46?

Also as for fuel levels, usually once the composite tank level(s) get to about 1/2 tank, the left side of the tank pumps down quickly due to the siphon pump and usually should be empty except due to hard cornering, then the tank pumps down very quickly again.

Use the Hiddern OBC Menu link in my signature to figure out if you are actually transferring fuel from the left side of the tank properly.
 
#465 · (Edited)
Yep, there is something like that, but I don't believe there is something to line up on the pump side. There is the sending unit, and the sucking jet pump, which keeps the fuel level balanced. I was adding that in to my post while you were writing this. It is all there now. :)

Don't you guys have work to do? I lead a life of luxury, and can respond all day, but shouldn't the rest of you be working? :)
 
#470 · (Edited)
Damn. Im 202k, 2nd owner (the first owner didnt replace the fuel pump), and as soon as I get my coolant system overhauled i'll be replacing my fuel pump. Its either luck, or using premium non-generic fuel and changing my filter regularly has extended the life of the pump. This might sound crazy, but before I change my filter I run a few gallons of E85 through my tank once its on Empty. Research has shown that it disolves some deposits left behind by the fossil fuel. The car runs fine on E85 up until around 4k rpm where the pump/injectors cant keep up and power slightly drops off. Im not looking for warnings and criticism for using ethanol, i'm fully aware of the risks and 'possible damage' that ethanol fuels can cause to non flexfuel vehicles. Just adding my 2¢.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top