I have spent some time dealing with this problem......
Here is my analysis of the errors in tank breathing system resulting in breaking fuel meters:
For your convenience the most important terms I use translated first:
Carbon Canister = Aktivkohlefilter
Tank Breather Valve = Tankentlüftungsventil
Four conditions are possible:
1. Good carbon canister AND Faulty tank breather valve (No danger)
2. (partially) clogged carbon canister and Faulty tank breather valve (very Dangerous!)
3. (partially) clogged carbon canister and Good tank breather valve (less Dangerous)
4. Good carbon canister and good tankbreather valve (normal situation)
I wil explain the system first,already been done here but for completeness.
The fuel tank needs to be able to breathe. When the fuel pump uses fuel from the tank, this wil result in low atmospheric pressure in the tank. Normally the fuel tank pressure is balanced with atmospheric pressure via the carbon canister.
(If the tank would be able to "breath" directly with outside air fuel fumes would enter environment non stop so that's why this breathing line is going through the carbon canister , the active carbon absorbes these fuel damps, the Americans call it so beautifully the "EVAP system" )
Also during hot summer temperature of the air in the fuel tank will rise resulting in increasing pressure and fuel damps wanting to escape the tank, these are also captured in the active charcoal from the carbon canister.
At some moment the carbon canister becomes saturated with fuel damps and these fuel fumes need to get out, but NOT in the environment. Thats were the second connection of the carbon canister comes into place. The engine intake manifold is connected to the carbon canister via the Tank breather valve.
Under certain conditons this tank breather valve opens and directly pulls air from the carbon canister (with all the fuel damps!) and lets it into the intake manifold to be burned in the engine.
The Tank breather valve is an oscillating valve, by switching an electromagnetic valve open and close in a high frequency the valve is able to regulate the airflow being able to run through the valve.
So far so good, we understand now how the carbon canister works as some kind of buffer in catching fuel damps. But of course the carbon canister must be able to breath as well !!!! And that is via the third hose connected to the carbon canister, a small plastic hose directly connected to outside air!
Situation 1:
GOOD WORKING carbon canister and FAULTY Tank breather valve.
During engine running the faulty tank breather valve can be temporary stuck in the full open position, temporary not oscillating. This cause the FULL engine vacuum pulling on the fuel tank via the carbon canister. A good working carbon canister has enough breathing airflow capacity to compensate the high negative pressure , the tank will not collapse.
Situaton 2: (the dangerous one!)
(partially) CLOGGED carbon canister and faulty tank breather valve.
The tank breather valve can get into a stuck open condition again and the full motor vacuum is pulling at the fuel tank via the carbon canister. Clearly the clogged carbon canister cannot compensate so a negative (temporary) negative pressure will exist in the fuel tank and collapsing the tank. Due to the very small clearance between tank wall en floating arm, the floating arm will crack the ceramic "platine" , and thats
the end of your fuel meterSituation 3: (a little less dangerous one)
Almost fully clogged carbon canister and good tank breather valve.
During long drives on hot days negative pressure can occur as a result of the fuelpump using fuel from the tank. It is very well possible the fuel pump is powerfull enough to create the necessary negative pressure to let it collapse. But is also very well possible the engine starts to stall before that and stop the tank from collapsing. This all depends of how bad the carbon cabister is clogged.
Situation 4 : all normal, no explanation required.
Another consideration is the carbon canister moslty gradually cloggs, but sometimes instantly after an overnight park when some insect has decided to set up a camp in the vent hose of the carbon canister. This vent hose it not protected by any means, it has and opening of at least 1 centimeter, unguarded wth thin wire for example.
If you go one step deeper you might ask yourself why no further damage is done to the fuel tank under (high) negative pressure then a small collapse. This i don't know for sure but my guess is the tank breather valve has a safety feature. Once stuck open and too much negative pressure is sensed it will close itself automaticly and cancelling the negative pressure at the other end of the line.
However i have read about a few cases were the complete fuel tank was so badly deformed exchange was needed, so this is all theory.
Based on the statistics from other worldwide Z3 forums and location of people interested in exchange sensor boards this problem is being seen the most BY FAR in the warmer countries with higher temperatures and relatively high (dessert) dust concentrations in the atmosphere.
In Europe for example the problem of the breaking fuel meter is being reported the most in France and Spain (Sahara dust during summers). Much much more then for example Germany, Netherlands, UK, other countries with lots of Z3 on the road.
This contributes to the theory dust and temperature play a large role in the speed of the carbon canister gradually clogging.
Another interesting fact. Good car mechanics (or BMW dealers) probably blow high pressure to the fuel vent line of the partially clogged carbon canister and the canister itself when a customer brings the Z3 in twice with a breaking fuel meter and then can conclude all is fine with the tank breathing system and the fault must be somewhere else (while it is not !)
But the real master (and i did this by accident
) will do this with a brand new carbon canister and a used one (from a problem Z) to compare and only then discovers there is actually a huge difference in the amount of air a new and used carbon canister is able to let through.........
Exactly this is the reason why I emphasize the words partially clogged in the first part of the analysis !
Just my 2c........