Thoughts on gas smell when engine first starts -- SOLVED!

[drtidmore]

I am well aware that the ECM pulses all cylinders with a couple of priming shots on start up and that may well be the source of the odor.  I also have the Well's Electronic's ICM which fires the engine within 1/3 revolution rather than the original design of 3 revs, so the engine fires in a very rich mixture environment. Regardless, when I first start up the Reatta and for a short time afterwards it smells like it is running REALLY rich!  I don't have any injector leakage as there is NO leak down of fuel rail pressure when I kill the engine (well at least none for 10-15 minutes).  The fuel rail regulator was replaced 3 1/2 years ago and it holds a vacuum so no ruptured diaphragm. The gas milage has not changed from its historic levels of 22-23 city mixed and 28-30 highway.  Once the engine runs for bit the odor vanishes.  I have gotten out immediately after starting and smelled the exhaust and the gas rich odor does NOT seem to be coming from it.  I have verified that the hoses to charcoal canister under the washer fluid tank are connected.  The source of the gas odor is mystery.

 

Any thoughts?

Edited May 6, 2017 by drtidmore (see edit history)

[D-a-n-i-e-l]

Check the purge canister and hoses going to it.

[drtidmore]

7 minutes ago, D-a-n-i-e-l said:

Check the purge canister and hoses going to it.

I have verified that all the hoses are connected on the canister as well as into the engine intake.  Is there a life on the actual canister (i.e. should it be replaced)?

[Ronnie]

My car has a smell of gas a lot of times when I first start it up. Always has. I figured it was a combination of the converter not being heated up to burn off any raw gas and the ECM not being in closed loop so the O2 isn't making optimum adjustments to the mixture. I never notice it inside with the windows up but if I'm outside the car or the windows are down I smell it. I think the direction the wind is blowing the exhaust from the tailpipe is a factor to.

[89RedDarkGrey]

The EVAP charcoal does have a "shelf life", just as with any component. I have never done it- but there are many videos explaining how to renew the charcoal, which does wear out- 28 years is way overdue. Your local aquatic pet store carries activated charcoal, and you can go the easier route by simply getting a "ready made" aquarium charcoal filter pack; it's a nonwoven material bag, with activated charcoal in it. EVAP canisters are usually ultrasonically welded polypropylene or ABS, very easily dissected then epoxied together.

 

Our Diagnostics AFAIK do not have purge solenoid continuity sensing (ES05 override to control) and an ECM code E045 Rich Exhaust would only appear if a quantity of extra fuel was present, but there is no EVAP leak sensing like newer cars have. A  NEW VALVE is cheap. 28 year old rubber diaphragms exposed to fuel fumes are easily brittle and hard.

 

You can also use a HANDHELD GAUGED VACUUM PUMP to check the system (valve, lines) too.

You just renewed the fuel tank- the EVAP is part of that system, too.

[drtidmore]

I thorough search for the evap canister turned up nothing other than the canister being discontinued.  It appears that it was pretty specific to the years our Reattas were built, but it was used on more than one GM vehicle during those years. I guess the only hope it to pull it, figure out a way to carefully crack it open and replace the charcoal.  Granted that does not sound like rocket science but it always is a gamble to crack open a part for which you don't have a suitable replacement on hand.  FYI, the part # is (GM part # 17085935)

 

FYI, the canister solenoid valve is "out of stock" at rock auto and it is simply not available at any FLAPS, so I question if it is another part that has become unobtanium?  I put a notify me on rock auto just in case they do get more in stock. 

Edited April 16, 2017 by drtidmore (see edit history)

[DAVES89]

Try Ebay. They will have discontinued stock on hand. That is how I found you NOS wheel speed sensor.

 I just checked there are 5 NOS units there from the same vendor $59.95 w/free shipping

Edited April 16, 2017 by DAVES89 (see edit history)

[drtidmore]

10 minutes ago, DAVES89 said:

Try Ebay. They will have discontinued stock on hand. That is how I found you NOS wheel speed sensor.

Dang, sometimes I get lost in the forest as I look for trees!  Found a NOS of the exact canister for a fair price no less (i.e. in line with what it listed for new).  Now I feel better about cracking open the original one and assuming success it will serve as my spare one.   Found BOTH the canister and the purge solenoid, so ordered both. 

Edited April 16, 2017 by drtidmore (see edit history)

[89RedDarkGrey]

52 minutes ago, drtidmore said:

Found BOTH the canister and the purge solenoid, so ordered both

 

Well that sounds great! Good idea, DAVES89  I was going to ask you to try and match one from HERE, I've never had mine out before, so I don't know what OEM looks like. It's basically just a 2 ported plastic box. BTW- sorry for that goose chase @RockAuto, I didn't notice the red lettering OUT OF STOCK  

[drtidmore]

8 minutes ago, 89RedDarkGrey said:

 

I was going to ask you to try and match one from HERE, I've never had mine out before, so I don't know what OEM looks like. It's basically just a 2 ported plastic box. BTW- sorry for that goose chase @RockAuto, I didn't notice the red lettering OUT OF STOCK  

89RDG, I tried that website but as I implied and they showed nothing. eBay totally slipped thru the cracks of my search.  I have purchased boat loads of stuff from eBay over the years, but this time, I was so focused on car parts vendors that I just let it slip my mind.  The canister for our Reattas is pretty specific as is the solenoid, (mechanically that is).  Yes, non of these are more than a tank of activated charcoal and a solenoid to open the canister to engine vacuum soon after starting to purge the canister of accumulated vapors from the gas tank, but given the tight confines of where Buick mounted the canister, it would take more research to determine if ANY of the slightly later model year canisters might work.  It would have been nice if GM had made the canisters renewable as some car makers chose. 

[89RedDarkGrey]

Thank you for not snapping at me, I'm always trying to help out Please- take some nice photos of stuff.

[DS_Porter 89 Burgandy]

  Just this week I had a strong smell of gas occasionally, usually when I shut down after pulling into the garage. A quick check revealed no leaks in or around the engine or underneath the car. It did not happen every time I drove it but when it did it was obvious outside the car. I checked the gas cap. I wiped it off and I was able to both blow out and also draw in. No good. So I cleaned the rubber seal and blew compressed air through and I believe I may have dislodged a piece of dirt because so far I have not smelled gas again.

  So before you do anything drastic like replacing the evaporation filter, clean or replace the gas cap. I believe the charcoal filter absorbs fuel vapor when the engine is off and then vacuum pulls it all into the intake and it actually draws it all out and the charcoal is once again ready for more fumes. I don't think it "wears out" although yes the valve could fail. It can be tested in diagnostics - override mode. The gas cap should be a one way check valve - air in but not out.

[89RedDarkGrey]

Yes, I agree- many times it can be as simple as a leaky tank cap. Activated charcoal used for this purpose (fumes) acts as a "sponge" for the V.O.C's and over time, does get saturated.There are many varied opinions on this topic. THIS is an excellent video, describing my earlier advice of repairing your existing canister. Some good explanations and advice can also be found HERE.

[drtidmore]

9 hours ago, 89RedDarkGrey said:

Yes, I agree- many times it can be as simple as a leaky tank cap. 

I checked the gas cap and it holds a fairly strong pressure before opening up and allows a suction with ease, but it is the factory original one, so I think I am going to replace it as it is just one more 29yo part that I can scratch off the list.

 

 

Edited April 17, 2017 by drtidmore (see edit history)

[ChrisWhewell]

I've had this gas smell also, haven't been able to figure it out. It would seem that the carbon canister is the only thing open to the air

[drtidmore]

I have the NOS canister and solenoid finally so next step is to replace the factory ones and see if it makes any difference

 

[89RedDarkGrey]

Use a dab of silicone paste or dielectric grease on the solenoid O-ring, it will slip in easier and ensure seal.

[drtidmore]

Well, I just finished up the installation of the new evap canister and solenoid.  A bit of a PITA as you have to remove the entire air cleaner assembly (top and bottom), the fuse/relay center, the washer fluid tank and the bracket on which the fuse/relay center sits (bolts on this were a bit tricky to find as they were somewhat hidden) in order to R&R the canister .  The fuse/relay center bracket also has mounted to its underside the headlight door control module which can remain attached to the bracket once the two connectors are removed.  Once everything is out of the way, the canister has a ring that holds it down (two bolts) and then the whole thing just slips out.  I removed and cleaned the bucket the canister sits in as there were some old leaves and stuff stuck to the outside of the old canister.  Took a couple of hours start to finish but at least no impossible to reach bolts or even a chance of a busted knuckle on this repair. 

 

The NOS parts matched to perfection the OEM right down to the numbers stamped on them and the molding marks in the plastic. The old canister solenoid was still holding a good vacuum when closed, but I have yet to investigate if it opened properly when energized.  I was expecting the old canister to smell like gas, but it had no odor even when I blew thru the inlet to force air thru the canister bottom, so don't know if this was a needed R&R or not, but again it was 29 years old, so just another item off the list.  

[89RedDarkGrey]

9 minutes ago, drtidmore said:

 I was expecting the old canister to smell like gas, but it had no odor even when I blew thru the inlet to force air thru the canister bottom

 

Did you try to blow air into the vapor line to tank, with cap removed? Hopefully that's not plugged.

[drtidmore]

2 minutes ago, 89RedDarkGrey said:

 

Did you try to blow air into the vapor line to tank, with cap removed? Hopefully that's not plugged.

I did not think to try that, but I will have another opportunity when I swap out the gas tank to the new one.  I will put it on my list.  Thanks.

[89RedDarkGrey]

33 minutes ago, drtidmore said:

when I swap out the gas tank to the new one

 

Don't accidently lodge a clog in your new EVAP system

[drtidmore]

I was thinking along the lines of gently blowing into the evap hose at the tank end.  If I can blow my breath thru the line without much effort the line should be fine

[drtidmore]

I may have found the source of the gas smell on startup and it could have been a disaster.  This morning, I went out and started up the Reatta to bring it up to temp and do a final addition of tranny fluid if necessary (once warmed up).  I opened the hood shortly after starting it and smelled gas which I had not smelled in few days and I did not smell while working on it yesterday.  I just happened to position myself such that I caught the glint of something wet on the front drivers side intake manifold.  I pulled the fuel rail cover to get a better look and damned if there was not a drip, pause, drip, pause, drip... of gas coming from the bottom of the fuel pressure regulator.  I noticed that the leak seemed to be diminishing as the engine heated which would also explain why the smell would go away.  I killed the engine first expecting that I had a leak on the return line fitting, but loosening and the resnugging that fitting did nothing.  I then pulled the regulator fully expecting to find it leaking at either the inlet joint or outlet (i.e. where the fittings attach to the regulator body), but absolutely NO leaks when I pulled a vacuum on each port (inlet, outlet & vacuum port).  FYI,the regulator had been replaced in sept '13.  If there was no leak internally on the regulator then it had to be a leak around the o-ring where the regulator joins the fuel rail.  I removed and inspected the o-ring carefully and could find no damage.  I inspected around the fuel rail and no cracks.  So I reinstalled the regulator and no leaks.  

 

What I am speculating (actually a WAG) is that the underside attachment bolt on the fuel regulator (PITA to reach at best) may not have been sufficiently tightened or worked itself slightly loose, preventing a solid seal at the o-ring.  That underside bolt is such a PITA to remove that I can't say for sure how tight it felt against the 1/4" racket as I removed it.  I know the forward bolt was VERY tight!  I WILL be keeping an eye on this until I am fully convinced that the leak has not returned on cold startup.  

[2seater]

That is a new one? Usually those o-ring connections are about bulletproof. Essentially it is the same as used on the injectors. Is this the oem Bosch regulator?

[drtidmore]

This is a Bosch regulator I installed about 3.5 years ago.  I know it was not a continuous leak as I have worked on the car enough times in recent months that I would have caught it, but I have been chasing a gas smell on first start.  It really appears that the o-ring seal was leaking when the engine was cold and would stop once warmed up.  As I mentioned, my gut makes me suspicious of the underside bolt on the regulator maybe not being sufficiently tight (that would be my fault), but there has not been ANY work on the rail connections that would have disturbed the 0-ring connection since that installation.  But if there was a slight under tightening of the lower bolt, the damned ethanol in the fuel may well have shrunk the rubber o-ring JUST enough to start a small leak under that condition.  As I said, I am going to keep my eye on it. 

Edited April 29, 2017 by drtidmore (see edit history)

[89RedDarkGrey]

1 hour ago, drtidmore said:

there has not been ANY work on the rail connections that would have disturbed the 0-ring connection since that installation.

 

I suspect- that inadvertently raising the rail pressure by in installing the new pump pointed out the leak that was nearly happening anyway. I always lube any O-ring with either silicone paste or Permatex green synthetic grease.

 

I'm very glad you caught this in time

[drtidmore]

2 hours ago, 89RedDarkGrey said:

 

I suspect- that inadvertently raising the rail pressure by in installing the new pump pointed out the leak that was nearly happening anyway. I always lube any O-ring with either silicone paste or Permatex green synthetic grease.

 

I'm very glad you caught this in time

The failing pump, when it was running was putting out 40 PSI which is in the same ball park as the new Bosch.  Could the new pump be related to the leak making its appearance, possibly, but the potential had to be there anyway as there can't be more than a couple of pounds of pressure difference if that between the two pumps.  Outside of my hypothesis of the lower bolt potentially not being snugged up completely, I can't explain why the leak occurred. 

[89RedDarkGrey]

Old pump had a pulsator, which calms down the pump spikes. The new turbine is constant- no sine wave at all. Relative pressures are same, but delivered differently. Maybe total bs theory?

[drtidmore]

1 hour ago, 89RedDarkGrey said:

Old pump had a pulsator, which calms down the pump spikes. The new turbine is constant- no sine wave at all. Relative pressures are same, but delivered differently. Maybe total bs theory?

Possibly, the new pump is obviously smoother as my fuel pressure gauge is rock steady where the old pump had some minor jitter.  I would think that the jitter would be more likely to pop a small leak than a constant pressure.  But I don't feel that the fuel pump was the root cause in any case, but rather the o-ring seal between the regulator and the fuel rail.  

[2seater]

Another casualty of the intake cough? I would think that might affect the diaphragm but your exam didn't show any fuel in the vacuum line? That might cause a momentary spike if fuel pressure as they will increase pressure under boost automatically. Maybe a small piece of debris under the o-ring. Injectors use the same system under the same pressure and you can spin them in place with no leaks? Just spitballin'.

Edited April 30, 2017 by 2seater (see edit history)

[drtidmore]

I had not thought about the regulator going to full pressure on the rail during the backfire, but I tested the rail at full pressure (i.e. using the green connector and the engine NOT running) after I installed the new fuel pump.  At that point, if there was an issue with the o-ring seal, it was not making itself apparent.  When I was getting ready to replace the fuel pump, I used the old pump to empty out the gas tank into 5 gallon gas containers and without thinking I did initially removed the return line from the regulator (as opposed to the feed line) and then I realized my mistake and reattached it.  If that back bolt was a tad loose, loosening and the retightening the return line  might have been enough to break the seal on the o-ring, but then again, I tested the fuel rail at full pressure of the new Bosch pump and I specifically ran my hand underneath the fuel pressure regulator to ensure that there was no sign of leakage.  I also did a fuel rail pressure leak down test and it held the full 40 PSI for well over 15 minutes, so that clearly says that there was no leakage at the regulator to rail connection. 

I just don't have a good explanation for this one.  Yesterday morning was a chilly one here (50) so I got out and fired up the engine to see if the cold might be a factor, but NO leaks all the thru the thermostat opening up completely.  Until my confidence in the regulator to rail seal is restored, I am going to keep a daily check for any signs of additional leakage.  

[2seater]

It isn't that the regulator goes to full pressure, it can go well beyond. The fuel pump is capable of 100psi or so, if deadheaded, and if you apply positive pressure to the regulator, the fuel pressure will rise in a direct 1:1 relationship. For example, when I was trying more turbo boost, with 10psi in the intake, I would regularly see 55psi on the fuel pressure.

[drtidmore]

2 minutes ago, 2seater said:

It isn't that the regulator goes to full pressure, it can go well beyond. The fuel pump is capable of 100psi or so, if deadheaded, and if you apply positive pressure to the regulator, the fuel pressure will rise in a direct 1:1 relationship. For example, when I was trying more turbo boost, with 10psi in the intake, I would regularly see 55psi on the fuel pressure.

Good to know.  I was not aware that the pressures could rise that high on the rail.  That being the case, such could be the causative factor in my situation.  

[drtidmore]

This morning for the first time since my last post I smelled gas fumes immediately after starting and sure enough there was a good drip, drip, drip coming from fuel rail pressure regulator.  At this point, I decided to pull the fuel rail itself so that I could get a good look at the regulator to rail joint...BINGO!  Apparently when I replaced the original regulator a few years back, the o-ring stuck to the inside of the fuel rail and I did not catch the problem.  I had NO leaks until recently  Once I saw the old o-ring jammed up against the top of the fuel rail opening, I got my dog legged awl and popped it out.  This was the problem as it was keeping the regulator from inserting all the way into the fuel rail and over time, this led to the leaking.  

[KDirk]

Well, cold temperatures did cause catastrophic failure of fuel o-rings on space shuttle Challenger once upon a time...good eye there to catch the leak visually. Lucky you found it, as that could've  had a nasty end result.  

[drtidmore]

I should make it clear that what I found was the OLD o-ring from the factory installed regulator lodged at the top of the fuel rail regulator fitting.  While the regulator did appear to insert into the fitting, the old o-ring was preventing it from seating properly. Yesterday, before I removed the fuel rail, after removing the regulator yet one more time, I used my cell cam to look up into the fitting to see if I could see a crack and what caught my eye was this green o-ring that appeared damaged.  That led me to remove the fuel rail so as to better inspect what I was seeing.  Once the old o-ring was removed the regulator seated into the fitting fully without ANY tension from the bolts which was very different that what I had been experiencing. 

[2seater]

Oops! I wish I could claim nothing like that ever happened to me, but I can feel my nose getting longer just thinking about saying it