Welcome to our resource for MG Car Information.



MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB Technical - Gasoline Smell in the Trunk (Boot)

fuel pump
C R Huff

In the trunk at the top of the fillerneck, there is a rubberhose. In my 78B it was leak.
the whole trunk smels fuel. after replacing the hose the smell was gone.

If you are getting fuel out the end of the pump it has a problem and could be a time bomb. On the other hand if you have not replaced the fuel tank or inspected the top of it, it may be the source of your smell, they are well known for rusting thru on top.
John H

Charlie - First of all, if the diaphragm is leaking into the interior of the pump, you are not going to get a fireball out of it. The vent on the cap of the pump has a check valve in it that prevents it from drawing fresh air into that area, thus the fumes will be such a rich mixture that it won't ignite. See the article, SU Fuel Pump Vents in the SU Fuel Pump Articles in my web site at:

I would check the filler hose on the tank as the most common source of fuel odor in the trunk area. If that doesn't cure the problem, then remove the pump and rotate it so the vent tube at the bottom of the coil housing points down and see if any fuel dribbles out at that point. If it does, then you will need to have your pump rebuilt. Cheers - Dave
David DuBois

Just had a similiar issue with a '77 mgb roadster that I've owned since July '07. As has occured with every B I've owned (four) the gas tank was suffering from pin hole rustout on its top and the gas vapors were entering the trunk. One new gas tank later problem solved!
Richard Miller

Thanks All,

I'll do a more thorough check of the plumbing and see if that shows any problems. Next Iíll pull the pump and turn it upside down. If I still havenít found it Iíll pull the tank and see if it is rusted on top. I havenít found rust anywhere else on the car, so I think I have some reason to hope it isnít the tank.

Iíll let you know what I find when I get back to it, but it is just one on the list of car jobs to do. Itís not one Iíve started driving yet, so it may be a while before I try to figure it out.

C R Huff

On my 74 midget, I was having a similar problem coupled with a tendancy for the gas pump not to shut off when the tank was full. It turned out to be the emissions hoses that I believe Willem referred to. They have this lovely metal braid on the outside that was in perfect nick, making it difficult to tell that the rubber inside was completely perished. I was not able to find a source for new hoses, so I carefully dissected the fittings at each end of the old hose with a dremel and cut-off wheel so that I could use jubilee clamps to attach a "normal" fuel line to the existing hose fittings. Easy to check; simply remove the hoses, plug one end with your finger and apply some air pressure to the other end.
David "no more fuel smell" Lieb
David Lieb


I will look for that as Willem suggested. I don't think I noticed a hose on the filler neck. Like I said, it's a new car to me. It does have bad hoses at the charcoal canister, but what I saw in the trunk looked like new.

C R Huff

I had the same problem with my 1973 B. Never saw a leak but the smell of gasoline in the trunk as well as inside the car when driving was always present.

I did two things. My B has a couple of braided hoses in the trunk. One goes from the top of the fuel tank to a vapor canister mounted inside the trunk near the right tail light and the other goes from the vapor canister all the way to a carbon filter located in the engine bay. These hoses are no longer available. I suspected that since the rubber hose is over 35 years old and subjected to fuel vapors ove that time span that perhaps in was micro cracked (or crazed) allowing vapors to pass through the hose and into the trunk. Since the hoses are braided I could not actually see the condition of the rubber. To test my theory I removed the hose from the top of the fuel tank and plugged the opening on top of the fuel tank. Make sure your gas cap is properly vented.

The second thing I did was to replace the braided flexible fuel hose at the fuel pump. If the hose inside the trunk was old and had microporosity then perhaps so did the fuel pump hose.

I am pleased to say that this solved my odor problem completely.

Good luck!
BOB in Atlanta
RES Schultz

Thanks Bob,

I haven't gotten back to check the car again yet. I work on the car about 12 miles from where I live. I will check that. Mine doesn't have the braided hoses. I don't know if that is because it is a 77 or because they have been changed. The hoses have a new look and still seem to have the "talcum powder" coating that some new rubber has. I suspect the previous owner tried unsuccessfully to solve the problem.

C R Huff

That talcum powder coating is called "bloom" and is caused by the cure agents in the rubber. The thicker the coating the longer the rubber has been in storage.
John H

The hoses Bob referenced are the hoses I was talking about. As Bob says, if you choose to eliminate these, you will need a vented gas cap or you will be starting a thread here about how your car runs fine for 5 miles or so then dies ;-)

If your hoses in the boot have already been changed, I am afraid that the most probable solution will involve a new gas tank.
David "not tanked" Lieb
David Lieb

Charlie. Most fuel pump related vapors would be outside of the trunk. It is the electro-magnetic coil portion of the fuel pump which sits inside the trunk, along with the associated electrics, while the fuel carrying portion sits forwards of the bulkhead. Very unlikely to be a fuel pump problem.

The vapor separator, the top of the fuel tank, the large rubber connection between the top of the fuel tank and the filler neck, and the hoses to the vapor separator are the most common failure points.

You need to remove the hoses going to the vapor separator and inspect them for cracking.

Check the tightness of the fitting that goes into the fuel tank and to which the line going to the vapor separator attaches. Tighten as necessary.

Attach a piece of fuel line (1/4" as I remember it) to to one attachment point of the vapor separator and the other end to a hand operated vacuum pump (if you have it). Plug the other line attachment with a finger and pump up the vacuum to about five psi. See if the tank will hold vacuum. If not, the tank is bad or you need to check your test procedure.

Check the hard line that goes forwards from the trunk to the charcoal cannister. It must be capable of passing air if it is to serve as a vent line for the fuel tank. I have seen them plugged before and "creative" methods of avoiding the difficult job of replacement. All allowed the car to run and all left a strong fuel smell in the trunk area.

Go through the system, piece by piece, from the pick up tube in the fuel tank, to the fuel pump, to the hard line going to the engine compartment. The fiber washers on the banjo connectors often leak and produce a fuel smell. The fuel pick up tube and sending unit have a rubber gasket between the fuel tank and the circular portion of the pick up tube assembly. This can go bad and I have seen it installed incorrectly. (Between the latch mechanism and the pick up assembly, which caused a strong smell of fuel, due to leakage, when the tank was filled.)

By the time you have done all of this, you should either have found the problem or have reached the conclusion you need to drop the fuel tank.

Les Bengtson

Well darn it. I thought that powdery look was a good thing. It did look supple with no cracking (at a glance).

I don't plan to eliminate the closed venting system. This is a 50,000-mile 3-owner car that is mostly unmolested. I say mostly because it has the downdraft 32/36 Weber and an exhaust header. It even has an aftermarket cat mounted farther downstream, though I wonder how long it will survive since the air pump has been removed.

I hope you are right that it is not the pump, and I also hope it is not the tank. It sort of seems to me that if the pump portion could leak into the coil section of the pump, I would see liquid fuel in the trunk.

Yes, I do have a Mity-Vac. Thanks for the reminder. I forgot about using it to isolate the problem.

I may get to diagnose it further this weekend. On the other had, if the weather is nice, I might just take the 1957 Lone Star to the lakes for a play day.

I will print all the tips and suggestions so that I can take them with me when I attempt the repair.

(sorry if this posted twice, the first time I tried it just stalled)

C R Huff

Charley - Don't run the pump too long with the end cover removed as it's secondary (or maybe primary) function is to hold the pivot pin at the back of the lower points toggle in place. Without the end cover in place, the pin will eventually work it's way out and the pump will cease to operate. I had to make up a couple of clips to hole the pin in place when I have a pump on my test stand so I can run them with the cover off. Cheers - Dave
David DuBois

Thanks David,

Ha, yup! In fact, I learned that one the hard way a few days ago. My first thought was, "I could have sworn I put that pin all the way in just a few minutes ago".

I haven't taken it out on the road yet, so it wouldn't have been too serious if it had died.

The new points and the cover are back on it now. Next I just need to track down the source of the fumes.

C R Huff

Well, I did go boating/swimming this last weekend instead of working on the MG. So, I'm just throwing this message in here to keep the thread in play so that when I solve it I can report back.

I know it is kind of disappointing when everyone helps to solve the problem and then never hears how it works out. So, maybe this coming weekend.

C R Huff

There has been some good comment with respect to this post but I would suggest that after 31 years there is a fair chance that water has splashed up and settled on top of the tank between the rubber strips separating the tank from the boot floor and caused pin hole rust right across the top face.
I have experienced this two or three times over the years and in each case replaced the tank - result - no further fuel odour either in the car or when opening the boot.
Peter M

You can generally tell if this is the case if there is staining down the sides/front/rear of the tank. Unless you never have more than a couple of gallons in the tank.
Paul Hunt

I was working on the car yesterday, but I got diverted from the fuel issue into the brake job. I figured if I needed to pull the pump or tank it would be easier to do on jack stands with the rear wheels removed anyway.

I was thinking like you, Paul, that if it were the tank I would see some signs. But, then I remembered that the car sat in a garage for a year and a half, and under double car covers outside for six months before I bought it. So, it might have rusted through while it was sitting. There is foam packing around the filler nipple that looks kind of gooey, so that may indicate the filler hose or the tank itself. I suppose it could also be age or a Coca Cola spill.

It might not be a bad idea to just go ahead and pull the tank. If it is not the problem, it could probably use a cleaning and painting to prevent the problem in the future.

C R Huff

Ha! I caught it red-handed this evening. It is the fuel pump.

I was letting it warm up in preparation to running a compression test. When I pulled the fuel pump wire in the trunk to bleed the carburetor dry before the test, I saw a little telltale trickle of fuel on the trunk floor. Previously it had only announced itself as fumes rather than liquid (there is no trunk carpet to hide the evidence).

Sometimes procrastination pays off. Iím glad I went on to other things instead of pulling the tank as I had contemplated. Yup, when the problem gets bad enough, the cause will become evident, eh? I can afford that luxury because this car has not been put into service yet.

In the spirit of thread hijacking, I guess itís okay if I hijack my own thread, though I may get fewer opinions than if I started over.

I read more compression than I expected. This is a 77 with an 18V engine, which has an ďLĒ in the engine number. Sorry, I havenít recorded the entire number. I was thinking the ďLĒ meant low compression, which I thought would crank at about 120 Ė 130 psi. I got 142 Ė 148 psi, warm with the throttle open, and I havenít adjusted the valves yet.

Also, what little paint remains on the engine is not black, but instead is orange or red.

Would others conclude that this engine has likely been out and has been fitted with the small dish pistons?

C R Huff

I just rid my boot of a faint but persistent petrol small.
I replaced my filler neck hose. with 4 inches of 54mm petrol hose., $7. (an American style filler neck).
The old rubber looked as new, however after I took it off and flexed it some tiny little cracks appeared. Another possibility is that if you have new hose, it might be too big. The guy in the shop initially gave me a 57mm bit.

Thanks Peter. If the fuel pump doesn't fix it, I'll look into that.

And, thanks to everyone that helped me diagnose this with the fewest steps.

C R Huff

This thread was discussed between 04/08/2008 and 18/08/2008

MG MGB Technical index

This thread is from the archive. The Live MG MGB Technical BBS is active now.