The summer is here, which means it’s top-down Jeep weather. But there’s a problem: My Jeep stinks like gas, and my neighbors are complaining, so I can no longer park it in my garage. An even bigger problem? I just changed the fuel pump, the fuel hoses, and the rubber grommets for the fuel vents. And it still smells like gas, but only in certain conditions. I can’t figure it out! So I need your help.
So, as any good online message-board-user looking for diagnostic help should do, I’ll begin by stating which vehicle I have and what problem I’m facing: It’s a 1991 Jeep Wrangler equipped with a 4.0-liter straight six bolted to an AX-15 five-speed manual transmission.
The issue is that the vehicle smells like gas, specifically when I let off the accelerator. It seems to smell worse with the top down, and when I’m on the accelerator driving steadily, it smells fine. But as soon as I let off the throttle to coast, I can just predict it: One Mississippi, Two Mississippi, thr — and there it is. Gasoline vapors have entered my nostrils.
This is peculiar; I’ve gone back and forth a million times on what this might be, and I’m still not sure what’s going on. To provide a bit of info on what I’ve done, here’s a photo of my fuel tank, which has a new fuel pump (and gasket), new fuel hoses supply/return hoses, new fuel vents, new fuel vent grommets, and new fuel vent hoses (not shown). The filler hoses are old, but appear to be in good shape.
The supply and return hoses plumb into hard lines, which seem to be in good shape:
There is a fuel filter in the supply line (part 22 in the diagram below), but it’s new, and in great shape.
The three hard lines go along the Jeep’s driver’s side frame rail, and then attach to flex hoses. One of the flex hoses (labeled 20 above) goes to a vapor canister like this:
The other two ports on the vapor canister go to the air filter housing and intake manifold. Like this:
As for the other two hard lines going from the fuel tank along the driver’s side frame rail, they go into a flex line like this:
Those flex lines have hard lines at their ends, and those hard lines plumb into a fuel rail. You can see the fuel rail and where the two flex lines plumb into it here in this photo:
And that’s it! That’s the entire fuel system. Very straightforward.
So what’s going on here? Why is it that, when I coast, the Jeep smells like gas? I have some thoughts.
One, it’s possible that when I’m on the gas, there’s lots of fuel flow going to the engine, and when I let off, that fuel then needs to be diverted back to the tank, since it’s not going into the motor. So maybe there’s an issue with the return line somewhere? I don’t see a leak, however.
I think the Jeep is running rich, and I have a check engine light that indicates this. And I think maybe that exhaust smell becomes so bad when I’m coasting that I can smell it from the driver’s seat. But why is it running rich? And why only when I let off the accelerator?
Could I have leaky injectors? The reason why I suspect this is that, when I coast in gear, the smell is worse than if I coast in neutral. In gear, the wheels are spinning up the engine, so there should be no fuel burning. But what if fuel is being sent into the engine anyway due to bad injectors. Could raw gas just be shooting out of my tailpipe?
There’s no liquid gas coming out of my tailpipe, for the record, and the exhaust stream looks normal. What’s more, injectors don’t tend to fail in groups of three, and I’m not sure just one single leaky injector would cause the Jeep to smell this bad.
Could my vapor canister be… full? Honestly, I’m not sure what’s going on. All I know is, the Jeep smells like gas when I let off the skinny pedal, and it’s precluding me from daily driving this thing in beautiful California, so I need to figure this out! Any help would be greatly appreciated.
Per the “he who smelt it dealt it” rule, I think it’s actually David Tracy who smells like gasoline off-throttle, not the Jeep.
The simplest answer is usually the correct one…
I’m going to go with: pinhole in the diaphragm of the fuel pressure regulator. This type of failure can cause a mysterious “too rich” issue, because you have fuel escaping through the diaphragm and into the manifold reference (vacuum) line which leads directly to the intake manifold. This could explain the notable fuel smell on closed throttle, because the low manifold pressure at closed throttle would more readily draw fuel through the pinhole AND the engine would have little need for fuel on closed throttle because that mode is typically a fuel cut (injector off) situation. This failure scenario could also explain the fuel smell when parked: after parking the car the fuel would bleed down through the fuel pressure regulator’s vacuum line, into the intake manifold, through the throttle body, and out through the air filter into the open air.
Here’s how you diagnose this: disconnect the vacuum line from the fuel pressure regulator shortly after shutting the engine off. Is there liquid fuel in the vacuum line? If so, you have found your problem. You can also disconnect and plug the vacuum line, start the engine, and watch for liquid fuel exiting the vacuum port on the fuel pressure regulator. Do not set yourself on fire when testing for this!
Great idea to check!
I forgot that this can happen with fuel pressure regulators… it’s bringing back a weird memory of it happening to someone’s car like 20 years ago.
Certainly a possibility.
Sounds like you just need to keep it floored all the time!
The “only when I let off the gas” bit points me to vacuum leak. Engine vacuum is high on deceleration.
I billionth the evap canister idea.
You can, in theory, blow these out and continue to use them. But, I’m sure a junkyard has a dry one already.
This is probably part of what was wrong, or led to, the fuel pump issues. I would also not rule out a clogged hard line. But, evap is first stop, IMHO. I would change the solenoid for good measure as well.
I’d guess the evap. canister is flooded with gas or the fuel return line to the tank is worse than it looks.
Make doubly sure you are looking at a Cali. spec. schematic and not a 49 state version.
Check for diagnostic codes with a scanner.
(Me awake tonight, staring at the ceiling)
My wife (to herself): He’s probably thinking of other women
Me (to myself): I HAVE TO KNOW WHY A STRANGERS JEEP SMELLS LIKE GAS
David’s not a stranger, he’s an honorary member of all our families.
I’m honored!
Oh, sorry, I was thinking of David Lee Roth. But you’re nice too!
Intrusive Thought Of The Day
It’s not that stupid 90 right before the fuel filter on the frame rail that changes sizes is it? I can’t remember if that’s a 4cyl only part or if it was on the 6’s as well.
I had a cracked canister in a Hyundai of mine that certainly led to a gas smell. I can’t recall it being more severe when letting off the gas though.
That canister was only accessible by dropping the fuel tank (Elantra Touring). No bueno. The purge valve is easier to replace on most cars, I would try that before replacing the canister. Outside of those EVAP related possibilities, Im not sure.
Time to get rid of it, it’s not an EV.
…/s
I agree that it is most likely either the evap system or a leaking injector or bad O-ring (if they’ve been removed). I can confirm that even one leaking injector can cause a whole lot of fuel consumption and a whole lot of fuel smell. I was once not careful enough in re-inserting injectors on my SE-R (I had let the car run on gas that was too old, clogging a couple of injectors, which I replaced with spares) and this led to torn O-rings that caused massive fuel consumption and pig-rich running, at one point being so bad that one cylinder was out of operation entirely due to the plug being saturated with gas, and raw fuel was going straight through and out the exhaust, and visibly leaking out of a small crack in the header. Smell was as you’d expect.
This was my guess as well. I had an 88 xj that I replaced the fuel rail on and tore the o ring on one injector installing it. That was my symptom, fuel smell during decel.
Even though this really makes me thing this is a charcoal canister / purge valve issue, your injectors are trivially easy to get to, and you can easily do a lame check / cleaning on the cheap… I’ve tested and cleaned injectors while broken down during a lemons race by putting just enough electrical tape on the straw of carb cleaner to make a good seal against the injector hole, hooked up a 9 volt battery to the terminals, and pressed down on the spray nozzle. As this will be at lower pressure than what the fuel rail supplies it’ll make any difference in spray pattern more obvious and you may even get to see one of them go from bad to OK as the cleaner works.
Its the evap system, either the canister is saturated, or the valve isn’t working.
I like the injector and evap canister theories.
If those don’t pan out, check the fuel tank filler neck for a crack. While there, check the seal on the cap. Unless I missed it, you mentioned replacing or inspecting everything else aft of the firewall, but you didn’t mention the filler neck or cap.
Love these types of stories. I’m hearing a click-click-click from my Dana 30, so I need to set aside a Saturday to check the front diff and bearings. In other words, you’re not alone.
Hoping it’s just a u-joint!
Me too!
No u
(Joint)
Logically, if it only happens when you let off the accelerator, you’re probably in a condition where there’s suddenly more liquid flowing through the line than is needed. Perhaps you’ve got a weak spot in a hose or a loose clamp joint and the water hammer effect is squeezing some out. Perhaps a pump stays on a bit longer than necessary, too. I’m not familiar enough with your system to know, so take all this with a grain of salt, but…
Way back when we had an old truck that was developing a split in a metal brake line, and we had a hard time tracking it down because as it turned out initially it only leaked when we applied the brakes while the engine was running. And when we were under it looking, we didn’t have the engine running.
Eventually it split enough for a “well, duh, that’s where the leak is” moment, but it took a while.
Leak in the carburetor float?
Oh 1991? Never mind. Oh wait, Grand Wagoners still had carbs in 91.
Oh not a Grand Wagoner?
Well something is making it go rich off throttle / high manifold vacuum.
Too down makes it worse? Maybe unburned gas out the tailpipe from the ignition timing going way off with high vacuum. Except it probably doesn’t have a vacuum advance distributor and it probably does have a catalytic converter.
You sure it’s a 1991?
I am afraid my diagnostic skills sort of poop out someplace in the late 1970s.
Is it possible you’ve just been spoiled by your fancy EV? 😉
his impossible little future-mobile with no noise, no emissions and the Giga-World interior
I love the idea that this is simple! However, I would replace the vapour thing. (shut up Nic fuel injection is witchcraft).
It could be a bad evap canister / purge valve (or the associated vacuum lines), but I’m thinking you may be on the right track with the sticky injectors.
The ’91 Cherokee with the HO 4.0 that my family had back in the day was the only car I’ve ever experienced where running fuel injector cleaner made a *significant* noticeable improvement in how the car ran. The 4.0 engine in the ’95 Grand Cherokee was totally insensitive to whether we ever fed it fuel injector cleaner or not. I don’t have a technical explanation for that behavior, but it was observable enough and obvious enough that I remember it 30+ years later, so that’s something…
Normally I would decry most of the products in the ‘rebuild in a bottle’ section of the auto parts store as snake oil, but in this case it is probably a good idea to run a can of fuel injector cleaner through the tank of your Jeep and see if the behavior changes. It’ll at least give you some additional information for your diagnosis.
Good luck!
How do your fuel trims look? The computer should tell you if trying to correct for injector leaks.
Jeep didn’t have OBD2 until 1996, so I doubt his ’91 OBD1 vehicle would spit out cylinder fuel trims… but I could easily be mistaken!
As others have mentioned, I’m leaning toward the vapor canister being faulty.
However, here is a longer shot: is it possible the new fuel pump has overwhelmed the existing pressure regulator (or the regulator has failed) and the injectors, etc. are receiving fuel at excessively high pressure all the time? Seems like that could exacerbate issues with aging injectors and could also introduce more fuel into the return-to-tank mechanism than it was designed to handle.
I’m leaning towards this hypothesis as well. If not this, then certainly something in the evaporative emissions system has failed.
The issue existed pre-new-fuel pump!
Hmmm. I suppose the regulator could have failed (or was marginal) prior to the new pump’s arrival. If it’s not regulating properly and is actually the culprit, the symptoms would have been there before and may have been made worse by the recent parts updates.
Most of the other relevant bits have been replaced; it might be worth replacing the regulator, especially if it’s the original item. OTOH I don’t picture regulators as a common point of failure. Hmm.
And on the assumption that it is the fuel vapor canister, a quick Google resulted in this page: 6 Signs Of A Bad Or Clogged Charcoal Canister (mechanicbase.com)
Based on what you wrote, you’ve got symptom 1 and 4.
I’m leaning this way…
Have you had any intermittent hard starting issues? One of the guys on OPPO has a similar era TC van whose issues were caused by excess pressures in his fuel system. The source? A clogged vapor canister.
Do you own a smoke testing machine? There are cheap ones available that will save you a lot of headaches tracking down anything vaccum, evap, etc.. related
Could the gas tank be so old it has some minor cracking and/or where the sending unit meets the top?
Could it be PCV related? Sometimes that can cause a gas odor
Like you mentioned, it could also be the charcoal canister. And it could be a compounded problem where you have 1-6 leaky injectors that have caused the canister to be too full, resulting in the smell. So it could be both issues.
Or, it could literally be that you need a new gas cap. That is also possible, and the easiest (and probably) cheapest part to throw at it.
Sorry, I keep hitting the edit button because I have more stream of consciousness thoughts about this. I have seen gas smell issues caused by intake gaskets as well.
I have a new gas cap!
Hmm, the CCV hoses hae been swapped as well. But the tank… sure, I bet there COULD be micro-cracks somewhere!
Believe it or not, there were plastic tanks on Dodge Trucks (and Ramchargers) back in the late 1970s… when basically no one was putting plastic tanks in cars/trucks yet.
The guys with those old tanks are all dealing with old plastic tank cracking issues. Plastic is better vs. metal tanks and do last longer (from my experience), but everything ages so….
I’d be willing to bet that it still smells when you’re accelerating or moving at a constant speed. Have you had anyone tail you to see if that’s the case?
I agree, deceleration seems like a red herring. The wind just stops blowing as strong as you decelerate. Can you apply some pressurized air to the tank and then search for a leak or source of vapor? Don’t know how hard it is to get to injectors but you can bench test them for correct spray and leakage.
I was wondering something related to this – if it’s in park and you rev the engine, does it smell?
This is the question I was hoping to find in the comments. How does it smell engine running while stationary. Ask Elsa (Eleanor? What’s the fake name of DT’s SO?) to rev the engine while standing watching under the hood. See what can be seen.
Elise!
It really is related to coasting! I’ve tried a million times; it only smells when I let off, even if for just a few seconds at a time, even if my speed doesn’t change much.
What if you do the same thing, but stationary?
Says who?
I’m not allowed to mention her anymore.
Boooo!!!
Most of us have been young and in love. We understand. Besides, we know your past. We are all still in shock a woman managed to clean you up.
It’s a damn miracle. I was hopelessly single for a decade!
Look at it this way. She Was smart enough to look under the dirt, motor oil, rust, and a jeep to find gold. Definitely a keeper.
David, I agree that the frequency of it is a bit weird. I can imagine how you must feel about ‘Elise’ – you probably feel some Esprit de corps with your partnered colleagues, and you might even feel a bit Elite. But it seems like you mention her Seven times a day, and by now Europa creek with your readers. Lotus have a break from her – you already Excel at what you do, and your Eclat will show itself more truly in your writing.
This is gold.
I love how your byline mentions you’re Jason’s friend. But how come his byline says “Also, David’s friend” as if it’s basically an afterthought that he only added because he didn’t want to look bad? Is this friendship not as mutual as we were led to believe??
I think David’s bio is terribly sweet and I’m not ashamed to say it makes me feel warm inside.