Author Topic: Hyperactive tank vent  (Read 84 times)

Offline Eayars

  • Curious
  • Posts: 5
Hyperactive tank vent
« on: September 27, 2026, 07:45:41 PM »
I’m starting to wonder what I got into here… Just picked up a 1990 K75S, my first K-bike. I can’t keep gas in it: fuel comes out the drain line at my right foot.

As part of the ‘new-to-me’ service, I replaced the fuel filter. While in there I checked the hoses and everything looks as it should, nothing loose or suspicious. The tank does not have a ‘burrito’, but the top couple cm of the tank appears to be a chamber of some sort that is probably serving the same purpose as the burrito. I just got back from an 85-mile shakedown ride: I started with a full tank and ended with the fuel warning light coming on just as I got home. I checked, and the fuel warning light is correct: there’s ‘slosh’ left in the tank and that’s about it.

Other possibly-useful information:
1) no, it’s not running rich. The plugs (changed them as part of the new-to-me service) look good, certainly not 17mpg rich. The gas is exiting at my right heel, not the tailpipe.
2) Yeah, it’s liquid gas, ml at a time, not vapor.
3) The tank has been sealed. There’s some sort of coating all over the inside, streaky in places. I’ve attached a photo taken inside the tank.

So… what’s up? I believe that the return overpressure line from the fuel rail comes back to the collection chamber, and the vent comes from the collection chamber: could the coating application have blocked drain holes to that chamber so that the fuel rail return is just pumping gas out the vent line? What else might be going on? Would it damage anything if I disconnected the fuel rail return hose inside the tank? What else could be causing this? Any suggestions? Are there supposed to be drain holes in this collection chamber, and if so how many and where are they? Maybe I could get a small drill bit in there and clear them. Really hoping to find a solution other than ‘new tank’.
  • Chico CA
  • 1990 K75S

Offline Laitch

  • Faster than a speeding pullet
  • Administrator
  • ^ Quintessential Motobricker
  • Posts: 11596
Re: Hyperactive tank vent
« Reply #1 on: September 27, 2026, 09:16:36 PM »
Return fuel comes from the fuel pressure regulator at a low pressure velocity and is dumped into the tank. Any blockage of return flow would show up as excessive fuel pressure, rich running and hard starting. Return flow in my '95 K75 is emptied from a tube on top of the burrito and splashes at low pressure into the fuel supply beneath.

The return flow from your Brick's fuel pressure regulator should come from a standpipe in the floor or drop freely from the ceiling plate. Which is it? You should be able to identify the location with the fuel cap assembly removed and the Brick idling.

Two ports exit from the floor of the tank—one dumps water from around the rim of the tank cap to prevent it from entering the tank. The other dumps fuel vapor—or excess fuel—if the tank has too much pressure or too much fuel within it. Both are often routed behind the right foot peg plate. Usually hoses are attached to those ports. Sometimes, they dump directly from their ports into a small plastic connector attached to the frame beneath the tank. This device is lyrically called an air accumulator in BMW's parts fiche.

Vapor needs to escape from the tank. It's up to the port in the burrito (maybe in the ceiling plate if your Brick has one) and the vent in the tank cap to keep tank pressure stable. You could remove the hose from the port in the tank's floor and use a wire to gently probe for blockage. To conform to CA's rules, BMW routed the vapor hose from the tank to a port in the crankcase. Attached somewhere in the run of the hose was a pressure valve that opened when tank pressure exceeded a certain value. The vapor would enter the crankcase then be transferred via the z-hose into the combustion air stream via the air box. It caused performance problems in states with more variation in temperature than California's and was rerouted as described in most other states.

If the fuel cap is opened when you stop after riding, can you hear air escaping with a whoosh?
  • Along the Ridley in Vermont.
  • 1995 K75 102,000 miles

Offline daveson

  • ^ Quintessential Motobricker
  • Posts: 1314
Re: Hyperactive tank vent
« Reply #2 on: September 27, 2026, 11:49:39 PM »
could the coating application have blocked drain holes to that chamber so that the fuel rail return is just pumping gas out the vent line?

Yep, that's on the cards. I'm thinking you have a burrito that's not shaped like a burrito, sometimes they look different.  There should be a small hole in the bottom so that any fuel that does inadvertently get in, can simply drain out back into the tank, rather than exiting the outlet drain. From memory I think there is also a hole at the top to allow excess fumes to enter the burrito. A photo might help.
  • Victoria, Australia
  • Current; '85 K100RT~100,000km; four other bricks. Past; Vulcan 1500, V Star 650, KLX 250(dirt bike) TT250(dirt bike)

Offline Eayars

  • Curious
  • Posts: 5
Re: Hyperactive tank vent
« Reply #3 on: September 28, 2026, 09:51:57 PM »
Return fuel comes from the fuel pressure regulator at a low pressure velocity and is dumped into the tank. Any blockage of return flow would show up as excessive fuel pressure, rich running and hard starting. Return flow in my '95 K75 is emptied from a tube on top of the burrito and splashes at low pressure into the fuel supply beneath.

The return flow from your Brick's fuel pressure regulator should come from a standpipe in the floor or drop freely from the ceiling plate. Which is it? You should be able to identify the location with the fuel cap assembly removed and the Brick idling.

Two ports exit from the floor of the tank—one dumps water from around the rim of the tank cap to prevent it from entering the tank. The other dumps fuel vapor—or excess fuel—if the tank has too much pressure or too much fuel within it. Both are often routed behind the right foot peg plate. Usually hoses are attached to those ports. Sometimes, they dump directly from their ports into a small plastic connector attached to the frame beneath the tank. This device is lyrically called an air accumulator in BMW's parts fiche.

Vapor needs to escape from the tank. It's up to the port in the burrito (maybe in the ceiling plate if your Brick has one) and the vent in the tank cap to keep tank pressure stable. You could remove the hose from the port in the tank's floor and use a wire to gently probe for blockage. To conform to CA's rules, BMW routed the vapor hose from the tank to a port in the crankcase. Attached somewhere in the run of the hose was a pressure valve that opened when tank pressure exceeded a certain value. The vapor would enter the crankcase then be transferred via the z-hose into the combustion air stream via the air box. It caused performance problems in states with more variation in temperature than California's and was rerouted as described in most other states.

If the fuel cap is opened when you stop after riding, can you hear air escaping with a whoosh?

Laitch, very helpful, thanks. I’ve attached a sketch of the tank innards on this beast. Instead of a ‘burrito’, the top 2cm of the tank (A) appears to be the ‘collector’. I can’t tell for sure, but it the edges of this collector are not smoothly joined to the top of the tank, and there may have been a gap between this and the tank top wall. The gap, if one existed, is full of tank sealer all around the edge. There is a black rubber hose from the pump to the filter, and a short hose from filter to a rigid pipe that exits the tank forward on the left side. (F) That exit goes via rubber pressure hose to the forward end of the fuel rail. The aft end of the fuel rail has a rubber hose that goes to a doodad behind the throttle bodies (the pressure regulator, presumably) and from there to another tank nipple farther forward on the left side. This left forward nipple enters the tank as a rigid pipe that goes to the top of the tank (bottom of the collector, actually) where it is welded to the bottom of the collector just forward of the filler, then bends down and forward. (B) From there, there is another flexible hose going to a nipple on the collector behind and left of the filler (C). Finally, there is a rigid line that goes from the collector just behind the filler, centered, that exits the tank at the back right. THIS IS WHERE THE GAS IS GOING. That nipple is attached to a line zip-tied down the frame to the right footpeg, where it puts out gas at a rate of roughly 4 liters/hour (average) while the engine is running. There is also a rigid line from the filler neck area through the tank and exiting behind this gas vent line, which is not part of this problem.

My best understanding of the problem is that as currently plumbed, the fuel returning from the rail goes to the collector chamber. Due to the enthusiastic application of sealant in this tank, there is no longer any exit from the collector chamber other than the vent line, so my fuel pump is pumping more gas out the vent line than into the combustion chambers.

My question is this: If I detach the line labeled “flexible hose 2” at point B and ziptie it onto the top of the return rigid line so it’s above fuel level, will that cause other problems? If I do this, then only vapor will reach the collector chamber, fuel from the regulator will dump at low pressure into the tank, and my ride won’t leave a smelly and expensive trail of fuel behind it everywhere it goes. I don’t see what problem it would cause, but SOMEBODY in Berlin once thought this was a good idea and I’d like to know their reasoning before I bypass what they did. What does the collector chamber do, anyway?

Daveson describes holes in the burrito, top and bottom, to let fuel that gets in get out. I think the collector in this tank had a gap around the top edge that served the same purpose, but that gap is filled now so gas pumped in from the regulator/return line can only exit via the vent hose.
  • Chico CA
  • 1990 K75S

Offline Laitch

  • Faster than a speeding pullet
  • Administrator
  • ^ Quintessential Motobricker
  • Posts: 11596
Re: Hyperactive tank vent
« Reply #4 on: September 28, 2026, 11:20:06 PM »
What's the answer to the "whoosh" question?
  • Along the Ridley in Vermont.
  • 1995 K75 102,000 miles

Offline Eayars

  • Curious
  • Posts: 5
Re: Hyperactive tank vent
« Reply #5 on: September 28, 2026, 11:24:02 PM »
No Whoosh.
  • Chico CA
  • 1990 K75S

Offline Ingo

  • ^ Motobrick Curious
  • Posts: 104
Re: Hyperactive tank vent
« Reply #6 on: September 28, 2026, 11:38:35 PM »
Tank pressure is pushing return fuel out, so no whoosh....
  • Janesville,  WI
  • 1992 K1100LT
You can't have everything; where would you put it?!

Offline Eayars

  • Curious
  • Posts: 5
Re: Hyperactive tank vent
« Reply #7 on: September 28, 2026, 11:41:05 PM »
Tank pressure is pushing return fuel out, so no whoosh....

Yeah, that's my take on it. I'm a bit curious how air is getting into the tank to replace lost fuel volume, but I guess there's an incomplete seal at the filler
  • Chico CA
  • 1990 K75S

Offline Eayars

  • Curious
  • Posts: 5
Re: Hyperactive tank vent
« Reply #8 on: September 28, 2026, 11:50:47 PM »
... I mean, even if this had the CA-spec vent to crankcase and everything was working as new, there would have to be some way for air to come into the tank to replace fuel as it was used. The filler cap allows that, right?

Dang, my /5 was a lot easier. That was just "oil goes in here, goes out there, forms a puddle under the bike, and don't try stopping because it has drum brakes front and rear."
  • Chico CA
  • 1990 K75S

Offline Laitch

  • Faster than a speeding pullet
  • Administrator
  • ^ Quintessential Motobricker
  • Posts: 11596
Re: Hyperactive tank vent
« Reply #9 on: Today at 12:14:20 AM »
BMW has made many changes to systems that have proven problematic during production runs. Tank remodeling has been one of them. Return line flow had at least three revisions included in three fuel tank versions.

You seem to be indicating that fuel flows across the ceiling baffle from B and down into D. If the upper end pipe of B is anchored somewhere, then the flexible hose attached to it can be rerouted to dump on the tank side closer to the pump and anchored by zip tie to a rigid pipe closer to the floor. That should end the fuel loss through pipe D.

It should be clear that with this type of tank baffle, filling past the flapper valve in the filler cap assembly will probably lead to some fuel loss to the pavement. The burrito models tend to avoid flowing leaks by means of the small size and location of the vent hole in the burrito. Your setup—once modified—will be suitable as long as there is no over-filling.

The purpose of the hose to the pump filter basket seemed to be the removal of air caused by fuel agitation when the fuel level became extremely low. Air bubbles drawn into the pump can damage the works. That kind of damage is called cavitation. The hose attached to the basket allowed air to rise up the hose away from the intake. A different system was designed and adopted around ’93. The fuel pump was changed from the 52mm with filter basket to the 42mm with a fuel filter sock, like many other brands adopted. The sock lessened or eliminated the possibility of cavitation.

Burrito envy is futile, of course, and so is the longing of things past, unless you start riding those things again.  :laughing4-giggles:
  • Along the Ridley in Vermont.
  • 1995 K75 102,000 miles

Offline Laitch

  • Faster than a speeding pullet
  • Administrator
  • ^ Quintessential Motobricker
  • Posts: 11596
Re: Hyperactive tank vent
« Reply #10 on: Today at 12:25:53 AM »
The filler cap does allow for air balance. It replaced a cap that caused fuel starvation.  :laughing1:
  • Along the Ridley in Vermont.
  • 1995 K75 102,000 miles

Offline Laitch

  • Faster than a speeding pullet
  • Administrator
  • ^ Quintessential Motobricker
  • Posts: 11596
Re: Hyperactive tank vent
« Reply #11 on: Today at 12:42:39 AM »
Here's a photo from a BMW service bulletin showing how the technician anchored the rerouted flexible fuel hose. That should work for you.
  • Along the Ridley in Vermont.
  • 1995 K75 102,000 miles

Tags: k75s fuel vent