I have experienced strong gas odors for years with a full tank + emission system issues and just came across this study plus this forum (very helpful). I bought a used CA car (a so called PZEV - partial zero emission vehicle) in TX and now live in NC. It's a 2007 E350 with 100k miles on it. When I called MBUSA Customer Service - and later the local dealer - I was told that my VIN does not show up on the Extended Warranty list so I'm out of luck. The Warranty Booklet says that for PZEV vehicles the Emission Performance/Control system is covered for 15 years/150,000 miles already so I'm guessing the folks I talked with don't know I am a PZEV. My question is whether anyone with a PZEV has dealt with MB and/or dealers and can provide me some coaching on how to get them to address my issues since my VIN is not showing up. I'll take the obvious step of waltzing down to the dealership with my paperwork but I'd be interested to know how others handled this issue. Thanks.
Analysis of Fuel Tank Leaks, Model Years 2003-2008
rds53benz
61 - 80 of 171 Posts
Daidnik, as reported many times in this post, the PZEV tanks are 100% steel. The lines are also steel with bolted connections.
Tabijan, you are the man. I searched that companies site and did not find the 8" option. That would involve drilling several bolt holes and may be a perfect bolt in. I have not seen how they hold the inner support in place but maybe an epoxy could solve that issue. It just needs to stay in place for assembly and the epoxy needs to be compatible with gasoline. JB Weld would work. Gas fumes break down any silicone type sealers.
Tabijan, you are the man. I searched that companies site and did not find the 8" option. That would involve drilling several bolt holes and may be a perfect bolt in. I have not seen how they hold the inner support in place but maybe an epoxy could solve that issue. It just needs to stay in place for assembly and the epoxy needs to be compatible with gasoline. JB Weld would work. Gas fumes break down any silicone type sealers.
Tabijan, I ordered the 8" and it appears it may work very nice. Fingers crossed. I had a plan to make a part similar to what they are selling. I had already chalked out the cuts I had to make in a mild steel plate. I was going to cut it out of 3/16 steel. The aluminum option looks like a better choice. Thank You. Now I just need to wait for parts. I am hoping I can get the template from the company so I can pre drill the tank holes. Then I could put the car back together and be ready when my parts finally arrive. Drop them into my new PZEV tank with a removable fuel pump! OK, that is what I hope to do. We will see what actually works out but waiting a week for a parts. I was ready to start cutting the steel plate but I had not solved the gasket issue yet, the 8" aluminum comes with a nice gasket and solves that problem also.
The factory pump ordered which I plan to take apart and remove the actual fuel pump for use in my module comes with a tank gasket. That might work to seal my opening. Still looking for all the answers. The submerged fuel thing is a concern. I have also looked at Orings but I don't have a machine shop to produce a proper sealing groove.
(EDIT 4-30-2022: There is a better pump to use, see post 110 for the updates. Mark)
I ordered in a Carter P76500M, Carter builds good electric pumps so I trusted this part. It is for the Non PZEV 2007 Mercedes so I know it meets the demand criteria of the car. When it arrived I took the entire module apart. Got the pump isolated and compared to the OEM Mercedes pump. They are very similar but not an exact match. I was able to do a few modifications and get it to fit quite well in the OEM module so it will go back into place like the OEM unit. I plan to test it before installing in the tank, just to make sure I don't install a pump that may not work. You have to test these pumps in gasoline or solvent so you have the flow through lubrication while it runs. If you start the pump dry it will damage the bushings in side in almost zero run time so never dry start one.
It looks like the pump replacement hurdle is now behind me. The next issue will be how to close the tank once it is installed. Sorry I did not take pictures as I worked to get the Carter pump properly located. There was a ton of trial and error, maybe not a ton but took me about two hours to get a fit I was happy with and yes, I modified a few things, not any functional things but minor things to get it to mount properly with this pump installed.
If I could have come up with the actual Mercedes Fuel pump it would have been a 10 minute install. I had to solder wires and modify the isolators that hold the pump in the module so it does not vibrate while running. Now just to solve the tank closure. Welding is the easy way but then if something goes wrong you are back to removing the entire undercarriage again and that is not a good option. The new cover plate has yet to arrive so I am on hold.
I ordered in a Carter P76500M, Carter builds good electric pumps so I trusted this part. It is for the Non PZEV 2007 Mercedes so I know it meets the demand criteria of the car. When it arrived I took the entire module apart. Got the pump isolated and compared to the OEM Mercedes pump. They are very similar but not an exact match. I was able to do a few modifications and get it to fit quite well in the OEM module so it will go back into place like the OEM unit. I plan to test it before installing in the tank, just to make sure I don't install a pump that may not work. You have to test these pumps in gasoline or solvent so you have the flow through lubrication while it runs. If you start the pump dry it will damage the bushings in side in almost zero run time so never dry start one.
It looks like the pump replacement hurdle is now behind me. The next issue will be how to close the tank once it is installed. Sorry I did not take pictures as I worked to get the Carter pump properly located. There was a ton of trial and error, maybe not a ton but took me about two hours to get a fit I was happy with and yes, I modified a few things, not any functional things but minor things to get it to mount properly with this pump installed.
If I could have come up with the actual Mercedes Fuel pump it would have been a 10 minute install. I had to solder wires and modify the isolators that hold the pump in the module so it does not vibrate while running. Now just to solve the tank closure. Welding is the easy way but then if something goes wrong you are back to removing the entire undercarriage again and that is not a good option. The new cover plate has yet to arrive so I am on hold.
1999 S500; W140.051; 2003 E320; W211.065; 1973 220; W115.010
Joined
·
1,286 Posts
JB Weld will work for a time, but will is not long term compatible with gasoline. It will eventually soften like a mud cookie with long term exposure to gasoline. I've used it and found out that way.
I have found only one epoxy with test data to support long term gasoline exposure. That is 3M EC-9323. 750 hr immersion in gasoline with almost no loss in matl strength. Great stuff, hard to get and expensive. Epoxy joint is a last resort. So many variables in trying to get a good glue joint.
There's really no way to do a leak-tight cover without some degree of custom machining. If it weren't for the submerged condition of those access holes when the tank is full, one could 'cheat' a bit.
The submerged condition changes this entirely because even the tiniest leak (far tinier than any EVAP leak check system will flag) will allow fluid seepage and gasoline really stinks badly so any leak will reveal itself in the full tank condition.
I have found only one epoxy with test data to support long term gasoline exposure. That is 3M EC-9323. 750 hr immersion in gasoline with almost no loss in matl strength. Great stuff, hard to get and expensive. Epoxy joint is a last resort. So many variables in trying to get a good glue joint.
There's really no way to do a leak-tight cover without some degree of custom machining. If it weren't for the submerged condition of those access holes when the tank is full, one could 'cheat' a bit.
The submerged condition changes this entirely because even the tiniest leak (far tinier than any EVAP leak check system will flag) will allow fluid seepage and gasoline really stinks badly so any leak will reveal itself in the full tank condition.
I ordered in some Viton Rubber gasket material. Viton is supposed to be resistant to gasoline for a couple dozen years. That should do it. It is approved for full time gasoline use. I will use a Viton seal and bolt the new cover plate in place. At least that is my current plan. The new cover comes with a Viton seal but it looks thin. The new material purchased is 1/8" thick and will seal it up well. If the aluminum new cover is not rigid enough for a quality seal my back up plan is to use that aluminum cover as a template and make one out of steel plate, either 1/8" or 3/16" thick for rigidity. I am feeling good that this plan will solve the tank issue. The Pump was tested today in the module and it worked flawlessly, tested in a 1 gallon bucket with about 1/4 gallon of gasoline in it, the pressure line was fed back into the bucket so it was a continuous flow situation. No problems with the Carter unit installed in the OEM Mercedes module. I hope to get the cover plate by tomorrow night but have other jobs to tackle this week. Might have time to install and re assemble by next weekend. Welding it closed is still on the table as a back up plan.
1999 S500; W140.051; 2003 E320; W211.065; 1973 220; W115.010
Joined
·
1,286 Posts
The Viton is fine; no problem with that. The difficulty is that getting a reliable, leak-free seal is very dependent on getting an even pressure/squeeze across the area of the sealing interface. If one of the sealing interfaces is thin/flimsy/not-true-to-form, then that is a problem. There's no easy way to grapple with the sheet metal surface of the tank. Both glue joint and elastomeric sheet seal are problematic due to flatness issues.
One could use a backing ring on the inside of the tank to sandwich the sheet metal between 2 beefier metal rings to force the sheet metal seal interface to better flatness, but then the custom parts become even more complex as there are more parts to make and the outer ring for the seal would need blind threaded holes on both sides. The threaded holes can't be thru holes or the fuel will just seep thru the threaded holes. The inner 'backing' ring would need to be split into 2 halves to get it thru the hole.
The beauty of welding the tank sheet metal seal interface is that visual inspection of the weld is a pretty reliable indicator of leaks and problem areas can be fixed with a 2nd pass.
The beauty of the o-ring seal is that all of the variables are determined by the geometry and the seal compatibility. The parts mate together metal-to-metal and your done.
One could use a backing ring on the inside of the tank to sandwich the sheet metal between 2 beefier metal rings to force the sheet metal seal interface to better flatness, but then the custom parts become even more complex as there are more parts to make and the outer ring for the seal would need blind threaded holes on both sides. The threaded holes can't be thru holes or the fuel will just seep thru the threaded holes. The inner 'backing' ring would need to be split into 2 halves to get it thru the hole.
The beauty of welding the tank sheet metal seal interface is that visual inspection of the weld is a pretty reliable indicator of leaks and problem areas can be fixed with a 2nd pass.
The beauty of the o-ring seal is that all of the variables are determined by the geometry and the seal compatibility. The parts mate together metal-to-metal and your done.
1999 S500; W140.051; 2003 E320; W211.065; 1973 220; W115.010
Joined
·
1,286 Posts
Yes I see now from the drawing how the backing ring sandwiches the Viton gasket between the fuel tank sheet metal and the cover plate with the 12X screws going thru the sandwich to threaded thru holes in the backing ring.
Items 3 & 6 on the dwg seem to have descriptions mixed up/reversed.
The 12X Nylon Spacers and button head screws do add a lot of seal interfaces/potential leaks; 24 places with no elastomer seals, only the squish of nylon.
I'm sure it will work. The only question is how well it will work. It will only affect that 3/4 full to 4/4 full state of the fuel level. In the worst event you won't want to fill up past about 3/4 full to avoid a puddle of fuel in that 'well'. The annoying part of that tank design is that there isn't even a way for the puddle to empty other than evaporation, thus the syndrome of fuel smell on the outside near the rear wheel.
Items 3 & 6 on the dwg seem to have descriptions mixed up/reversed.
The 12X Nylon Spacers and button head screws do add a lot of seal interfaces/potential leaks; 24 places with no elastomer seals, only the squish of nylon.
I'm sure it will work. The only question is how well it will work. It will only affect that 3/4 full to 4/4 full state of the fuel level. In the worst event you won't want to fill up past about 3/4 full to avoid a puddle of fuel in that 'well'. The annoying part of that tank design is that there isn't even a way for the puddle to empty other than evaporation, thus the syndrome of fuel smell on the outside near the rear wheel.
I look at the opposite side where Mercedes mounted the filter, same exact molding and fuel exposure. Left side and Right side of these tanks look identical. The fuel filter door seals and does not have leak issues so I think if this gets sealed with the Viton and passes a pressure test I will call it god and install it. I will only pressure test to 10-15 PSI in the tank since fuel vapor will never expose the tank to more than that and much more will cause the tank to expand like a balloon. Still not parts to install the new cover. UPS says it won't be here till Tuesday PM now. Dang, I had tomorrow pretty clear to start work. Wednesday is booked so on hold for a bit.
1999 S500; W140.051; 2003 E320; W211.065; 1973 220; W115.010
Joined
·
1,286 Posts
Yeah I wondered about those studs. They look like they might be PEM studs in sheet metal, but how they are fluid tight is a mystery to me.
I hope it all works well. You're in some uncharted territory but why not give it a shot?
Be careful with internal pressure testing on the fuel tank. They can't support much pressure differential inside to outside. You have to look at it as force per unit area and even at 1 psi, the tank has so much area that there is hundreds of pounds of force pushing outward to 'balloon' the tank and the tank can easily get distorted in unexpected ways. I've found that out by direct experience on smaller and simpler tanks than that one.
When factories leak check they have to use external bracing on the large areas to pin the tank in on both sides so it won't distort.
There may be some pressure regulators to work at settings around 1 PSI, but the ones I've used don't regulate very well at low settings and tend to creep upwards even at a few PSI setting.
I hope it all works well. You're in some uncharted territory but why not give it a shot?
Be careful with internal pressure testing on the fuel tank. They can't support much pressure differential inside to outside. You have to look at it as force per unit area and even at 1 psi, the tank has so much area that there is hundreds of pounds of force pushing outward to 'balloon' the tank and the tank can easily get distorted in unexpected ways. I've found that out by direct experience on smaller and simpler tanks than that one.
When factories leak check they have to use external bracing on the large areas to pin the tank in on both sides so it won't distort.
There may be some pressure regulators to work at settings around 1 PSI, but the ones I've used don't regulate very well at low settings and tend to creep upwards even at a few PSI setting.
I ruined a tank once by trying to use 30 PSI. Did not explode just expanded till it cracked, dumb mistake and I had to install a new tank. Those lessons don't leave you once learn the hard way. My pressure regulator is a nice one, I can dial back to any pressure I want. I will go less than 15, maybe try 1PSI and work up to 5 PSI and stop. Wait and leak check. The tank is Steel so I don't expect this one to fail. The one I ruined was in a boat and was fiberglass but it still did not have to be ruined. I was just draining old fuel from the tank using Air pressure to push the fuel out of the tank when it cracked. I guess more air went inside than gas was exiting. Thanks for the ongoing consultations. This tank can be fixed and reused but I want to be clear it is not for a first time mechanic to try. There are a lot of obstacles to work out fixing the OEM tank with a new fuel pump. These are nice cars and should be repaired not scrapped. The Dealership option of $2400 for the tank and 8-12 hours labor at $150 per our adds up to a $4,000 shop visit for a fuel pump failure. A $100 part. At OEM Prices maybe a $250 part but still, not reasonable to make this a $4,000 job and why did they keep selling this exact tank without the fuel door in it. That to me is just silly. They want you to roll over and buy a new car. Inside the OEM Tank the pump is set up with quick disconnect locks that hold it in place. No joke, once the tank is open the pump is out in less than 5 minutes. With an access door this would be a 1 hour job in any garage. It would also be nice if the actually sold the OEM Fuel Pump module that is inside this tank. Both would make this a very easy job.
Not much to update. The cover came in, at 8" it was a little to big, I cut it down to 7 7/8" and it fit better the factory tank has a round area that is flat above the fuel pump, this round is just under 8" and then it curves up to a higher level, this curve was keeping my cover from laying down flat. Took quite a while to accurately drill all the holes, 18 I think. 16 for the cover and 2 that hold the back up plate in place. Once that was done the install is straight forward with the tank out of the car. Got everything assembled and installed the repaired tank back into the car. No Joy, only 20 PSI at the injector rail? Started checking everything, everything checks out pretty well. Kept looking for the cause. Went to pull the new cover off in the car to look at the pump for problems and discovered the new cover is too large to remove in the car. 7" opening with a nearly 8" cover. I was able to slide it over on top of the tank and below the body which gave full access to remove the fuel pump module or check pump operation. My scan tool has a feature to activate the fuel pump and once powered on I found my leak quickly. The pressure hose runs right on top of the tank. It had a small cut in it where I had cut the opening for removing the pump. I cut the tank with a small dremel tool and did not let it go deeper than maybe 1/8", this hose had to be in the perfect spot to be damaged but that is my luck. Did not see it at tear down dang it. Would have been much easier to fix while the tank was clean and sitting on the ground. So I removed everything once again. I had not yet installed the driveline or exhaust so the job was a little less this time. Cleaned the tank once again. Did a lot of reading and study on what type hose can be used inside the tank. There are hoses available that are a type of rubber. That was my number one choice because of the flexibility and ease of installation but I can't get product for several days. Not one parts house I called stock the submersible product. I did read that many motorcycle shops have to use this hose but did not find any with 3/8 inventory. I ordered my replacement hose in Nylon fuel hose Dorman brand, 15 foot for $20 It should work with plenty left to practice on for a good fit, it is not as flexible as the factory Nylon hose but it does not need to be really flexible to work in this application. The factory hose with the cut in it will stay in place. It is tied down inside the tank and there is no access to remove the hose. There seems to be plenty of room to run a replacement hose. Automotive type fuel hose will fail if submerged in fuel. The spec for hose that works submerged is SAE 30R10. There is also a 30R9, 8,7,6 Etc. Non of these will work, only the 30R10. This is a flexible rubber type hose that looks like traditional fuel hose but it costs about $15 per foot or more depending on where you find it and it is rated at 100 PSI. The E350 only has 61 PSI so it would work fine. The in tank hose is a 3/8" hose. Gates and Dayco both make the 30R10 hose. You can buy it in 1' or 15' lengths, the 15' length is about $165. Gates 27081and Dayco 80161, it is so expensive they sell it by the inch at some places. I think 36" would be enough. I wanted this type hose but ordered the Nylon hose. Nylon meets all the specs, is pressure rated much higher than the 30R10 the only negative is it is harder to work with. Special training to install fittings although you can buy quick connect fittings for the Nylon connections. I plan to re use the Mercedes connectors and install new hose. I will check back in once this is done. I need to be away from the project for a week starting Tuesday so if parts don't come in there may be a delay.
If I had a Do Over I would cut the opening smaller, my opening cut was 6", 5" would be enough. ( and maybe avoid cutting the pressure hose) The pump module measures 4.5". The new cover set up does seal pretty well. It is held in place by 16 bolts, with only 4 in place and with 10 gallons of gas in this tank I pulled it out of the car and it did not spill at all. I installed it with just 4 bolts to avoid making a mess while removing the tank, it worked. Nylon fuel line seems to be the line of choice at the factory in most cars today, it is not expensive and lasts a very long time. The OEM hoses are designed to flex, the aftermarket hoses will flex a little but not nearly as well as the factory hose.
Now that this is nearly done I have confidence the repair is a good reliable repair, it has been a lot of work but most of the work would have to be done to remove and replace the fuel tank regardless so the time spent repairing the fuel pump is just a cost saver and it sets you up to have easy access should the pump ever go bad in the future. Or in my case when you find one of those nylon fuel lines is leaking you can get in there for repair! Glad I did not weld it shut.
If I had a Do Over I would cut the opening smaller, my opening cut was 6", 5" would be enough. ( and maybe avoid cutting the pressure hose) The pump module measures 4.5". The new cover set up does seal pretty well. It is held in place by 16 bolts, with only 4 in place and with 10 gallons of gas in this tank I pulled it out of the car and it did not spill at all. I installed it with just 4 bolts to avoid making a mess while removing the tank, it worked. Nylon fuel line seems to be the line of choice at the factory in most cars today, it is not expensive and lasts a very long time. The OEM hoses are designed to flex, the aftermarket hoses will flex a little but not nearly as well as the factory hose.
Now that this is nearly done I have confidence the repair is a good reliable repair, it has been a lot of work but most of the work would have to be done to remove and replace the fuel tank regardless so the time spent repairing the fuel pump is just a cost saver and it sets you up to have easy access should the pump ever go bad in the future. Or in my case when you find one of those nylon fuel lines is leaking you can get in there for repair! Glad I did not weld it shut.
Places like Oreilly always have the nylon Dorman hose in stock.
61 - 80 of 171 Posts
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
-
?
