Tracked a fuel smell to the EVAP canister, but why is it there?

[TrevorS]

I've been noticing a scent of gasoline (not strong, but distinct) in the vicinity of the fuel filter for awhile now (no idea when it first appeared, or if it might even have always been there). I checked all the under-hood fuel path connections both engine idling and off and all is good.

Yesterday I pulled the battery tray, checked the EVAP connections (seem fine) and removed the canister, the scent is very strong from the underneath breather.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

As a final test of the fuel system, I plugged the check valve outlet where it leads to the canister to make sure the scent ceased. Opened up the hood today and sure enough, no scent of fuel anywhere. The plug I'd put in didn't want to let go and the pipe opening made a momentary sucking sound when it finally did, and then I could hear a very low popping sound (which grew gradually quieter until it stopped) as though equalization was somewhat constricted. Guess the cool overnight temp caused the tank vapor to contract creating a slight vacuum.

Thing is, I understood the check valve has two jobs, one being to slightly resist vapor from the tank to allow mild pressure to build, the other to prevent the return of vapors from the canister to the tank. I'm wondering if I may be getting excessive venting from the tank resulting in charcoal saturation and perhaps also a problem with leakage back to the tank. I looked in the manual for a procedure to test the check valve but can't find anything. I'm thinking I should try cleaning it.

I'm also wondering if an associated problem could be due to driving conditions short of what the ECU requires for purge. The car is mostly parked outside with a fabric cover and some tree shelter, only occasionally driven and at speeds typically little over 25mph -- don't now exactly what the ECU needs to see before it purges. (The purge vacuum hoses are fine, as are manifold vacuum and engine behavior.)

Anybody have experience with this?

========================LATER=====================

Thing is, I understood the check valve has two jobs, one being to slightly resist vapor from the tank to allow mild pressure to build, the other to prevent the return of vapors from the canister to the tank. I'm wondering if I may be getting excessive venting from the tank resulting in charcoal saturation and perhaps also a problem with leakage back to the tank. I looked in the manual for a procedure to test the check valve but can't find anything. I'm thinking I should try cleaning it.

Decided to take another whiff of the canister 24 hrs later and guess what, no scent of gasoline anymore! So, just sitting upside down overnight was apparently enough for the vapor to dissipate and it clearly didn't contain any liquid gasoline when removed. Removed the check valve, cleaned the exterior so I could see how it was supposed to function, then tried blowing through it in both directions with the check balls seemingly clear of the ports. Couldn't push any air through and so poured a little gasoline into both hose ends and started shaking. Turned out it was at least partially blocked and may not have been properly closing in either direction. Hit it with two gasoline passes and lots of shaking and am hoping that solves the problem.

PS. Got the EVAP system re-installed, but can't go any further until I derust and repaint the battery tray. Still, I can check under the hood occasionally and see if there's an improvement -- it's definitely better right now, but kind of early to tell.

 

[TrevorS]

Working on the tray derust, but noticed that (no big surprise) there's again a slight scent of gasoline in the area of the fuel (EDIT: filter not pump). Now, this could either be entirely normal, or it could mean my EVAP check valve still isn't sealing properly -- at least I'm certain it isn't the canister or the purge controls.

Problem is, nobody at all is chiming in to tell me whether or not they also detect a slight scent of gasoline in this area. Surely you haven't all removed your evaporation system components? Not all states allow you to do that and still pass emissions. Can anyone chime in on this?

 

[PieEyedPiper]

Working on the tray derust, but noticed that (no big surprise) there's again a slight scent of gasoline in the area of the fuel pump. Now, this could either be entirely normal, or it could mean my EVAP check valve still isn't sealing properly -- at least I'm certain it isn't the canister or the purge controls.

Problem is, nobody at all is chiming in to tell me whether or not they also detect a slight scent of gasoline in this area. Surely you haven't all removed your evaporation system components? Not all states allow you to do that and still pass emissions. Can anyone chime in on this?

This is all I'll be good for one this one; I do not have a fuel smell in engine bay.

 

[TrevorS]

This is all I'll be good for one this one; I do not have a fuel smell in engine bay.

It's only a slight scent and I notice it only above the charcoal canister which is under the battery tray and adjacent the fuel filter in my 1990. I can't detect it anywhere else under the hood.

 

[estoyaqui30]

I'm sure if this is relevant because I have a 2G, but I have no gas sell at all. Did you check the canister for any leaks, just some slight crack maybe. I would assume, that the 1G is the same as the 2G, in that the canister holds the vapors till they are sucked back into the intake at the appropriate time.

 

[TrevorS]

I'm sure if this is relevant because I have a 2G, but I have no gas sell at all. Did you check the canister for any leaks, just some slight crack maybe. I would assume, that the 1G is the same as the 2G, in that the canister holds the vapors till they are sucked back into the intake at the appropriate time.

With the canister removed, the gas smell is only noticeable at the underside of the canister (the breather), nowhere else, and the exterior appears perfect. I don't know your canister location nor whether it has a breather. Yours is OBDII and so it's very possible your canister is completely closed and uses a solenoid valve to control air flow in. Unlike mine, that would fully close the EVAP system.

This is all I'll be good for one this one; I do not have a fuel smell in engine bay.

Just realized yours is also 2G, which is again OBDII and quite possibly a closed canister. I wouldn't expect an OBDII car to have a non sealed EVAP system, whereas requirements were a little looser in 1G days. Would be interesting to know whether the 2G canister has a breather.

 

[forcefed_90tsi]

Check for any type of small leaks. With the canister located under the battery tray, it may have gotten some battery acid on it somewhere.

 

[PieEyedPiper]

Would be interesting to know whether the 2G canister has a breather.

.. it does now! Some years ago I smacked a rock in a parking spot. There was no damage apart from having to dislodge a 6lb rock from the opposite inlet of my side mount. On a 98, that's where the charcaol canister sits. I have a large split in the container that it likely leaking like a mofo but it only started giving me a P0144 I think? It's "evap gross leak detected". Fuel mileage has never been better so I just clear the code once every 2 months or so when it pops up.

If I were completely fair I'd admit I've made a few assumptions about that scenario but so far so good.

Enough about me. Have you tried pressure testing the canister?

 

[TrevorS]

.Enough about me. Have you tried pressure testing the canister?

Pressure testing would accomplish nothing since the '90 EVAP tank has a large breather (see above photo), it isn't intended to hold pressure.

Moving on to today, after some 28 hrs of forced air flow through the canister inlet (venting through the outlet and breather) and a considerable drop in the noxiousness of it's venting (I could swear it feels lighter, measured 2 lbs weight, but unfortunately, didn't weigh it earlier !), I put the EVAP system back together and did some basic vacuum checks (hunted online and finally bought an "OEM" branded Mityvac 8000 from a relatively local Autozone). Everything seemed to work back to the purge solenoid, but then I hit complications.

The solenoid is vacuum open (no voltage applied) and so I tried measuring the coil resistance, came up with around 3.8Kohms using both digital and analog meters. As a reference, I tried measuring the impedance of the fuel pressure regulator solenoid adjacent it and couldn't come up with any number at all (behaved like an open circuit). After reinstalling everything I'd removed, I took it for a run and it behaved entirely normal (at least in my experience). From reading, it appears the fuel pressure regulator solenoid is responsible for protecting against fuel line vapor during warm engine start, and the EVAP solenoid is responsible for EVAP canister purge. My impedance measurements suggest neither solenoid is operational -- am I missing something ? Could they have been killed when my original ECU died and a power pack was connected to the car to attempt restart?

 

[PieEyedPiper]

Pressure testing would accomplish nothing since the '90 EVAP tank has a large breather (see above photo), it isn't intended to hold pressure.

What I meant was for you to check if it's exiting the canister by other means than the filter, which might produce the unpleasant scent of gasoline. Sounds like you know what you're doing.

 

[H@xtGSX]

The solinoids could have been fried, though I don't know how much that would affect the smell. I have no emissions stuff, but prior to removal, I found that none of the stuff controlling my evap system was working anyhow and I never noticed a gas smell. Try bypassing the canister and see if it still smells... If it does, I would start looking at the banjo bolts on the fuel filter and the hose for the gas tank vent line, since they are in the same area.

 

[TrevorS]

What I meant was for you to check if it's exiting the canister by other means than the filter, which might produce the unpleasant scent of gasoline. Sounds like you know what you're doing.

The solinoids could have been fried, though I don't know how much that would affect the smell. I have no emissions stuff, but prior to removal, I found that none of the stuff controlling my evap system was working anyhow and I never noticed a gas smell. Try bypassing the canister and see if it still smells... If it does, I would start looking at the banjo bolts on the fuel filter and the hose for the gas tank vent line, since they are in the same area.

Thanks guys !

I've gone over the canister with the fabled fine tooth comb, it's not leaking, just venting out the breather. I also checked the purge valve and the hoses on both sides of the canister and all's tight. The fuel system is also tight, verified no smells anywhere with the canister removed and check valve outlet line blocked. (I think the '95 is also OBDII, in which case the canister was probably sealed.) Guess I've been banging my head against this thing for a couple weeks now, but picking up that hand vacuum pump made a huge difference, finally able to do some component testing.

Turns out the fuel pressure regulator solenoid is stuck in the position that ties the regulator to the plenum and since I don't notice any problems with warm engine start, I guess it never gets hot enough around here for the fuel in the rail to vaporize. I've tied the two vacuum lines together bypassing the apparently dead solenoid -- will see if I notice any differences in behavior. I tested the EVAP hose to the intake and it holds vacuum regardless of engine conditions, whereas with a warm engine at 3K rpm, I read the purge valve should open -- it doesn't. So, I've ordered a new EVAP purge solenoid, but since it's coming from Atlanta, guess it'll be a few days.

In any case, it appears the reason there's a scent of fuel in the vicinity of the fuel filter is that the canister had become saturated thanks to never being purged. When the new solenoid comes in, I'll have to repeat that last test and see if I detect a little purge action. Sure hope I do.

 

[TrevorS]

Turns out the fuel pressure regulator solenoid is stuck in the position that ties the regulator to the plenum and since I don't notice any problems with warm engine start, I guess it never gets hot enough around here for the fuel in the rail to vaporize. I've tied the two vacuum lines together bypassing the apparently dead solenoid -- will see if I notice any differences in behavior. I tested the EVAP hose to the intake and it holds vacuum regardless of engine conditions, whereas with a warm engine at 3K rpm, I read the purge valve should open -- it doesn't. So, I've ordered a new EVAP purge solenoid, but since it's coming from Atlanta, guess it'll be a few days.

In any case, it appears the reason there's a scent of fuel in the vicinity of the fuel filter is that the canister had become saturated thanks to never being purged. When the new solenoid comes in, I'll have to repeat that last test and see if I detect a little purge action. Sure hope I do.

Picked up the new purge solenoid today and measured the coil resistance, roughly 41 ohms -- fine! Swapped it in for the original, reran the purge valve to intake vacuum test, and with a well warmed engine and idle above 3K, the canister purged -- Yay!

However, I'm now noticing a scent of gas at the bottom front of the open driver's door. Just under there is where the fuel and vapor lines enter the engine compartment, so looks like something else needs tracking down .

 

[RSTurboFWD90]

I was able to replace the charcoal in the OEM canister. I posted photos here.
https://plus.google.com/u/0/photos/102798776259789403620/albums/5898351068819664945?partnerid=gplp0

You must be logged in to view this image or video.

 

[Bud92gsx]

I removed all my emissione.No fuel smell in the engine bay.

 

[VisegripMech]

Haven't re-read the fine details in the post but did have a similar problem and it turned out to be the stock fpr leaking fuel into the vacuum lines.

 

[RSTurboFWD90]

Good to keep in mind. In my case, I would smell gas only after shutting the car off. No odor while running/driving and no performance issues. I checked all the other EVAP system components and they all seemed to be working properly. I removed the charcoal from the canister and it smelled strongly of gas odor. Even sitting for days in open air did not reduce the odor level. I have now put the rebuilt canister back on the car. A short test drive produced no odor. I will drive it more this weekend. I hope the problem was the saturated charcoal. If the odor returns, then I will have to dig in again. I thought this rebuild may be useful to others since I could not find any aftermarket canisters for this car and OEM is no longer available. I have a completely stock car and wanted to keep it that way.