Bad fuel mileage

[upnsmoke]

Well, it's raining again, so I'll go ahead and write up this post. I'm working on a 99 Ex manual trans with 80k miles. The car has no drivability issues, idles smoothly at normal speeds, and pulls strongly as it should. The problem is that the fuel mileage is currently 24 mpg in the city, and has been as low as 19. It is running rich enough to get the oil smelling like gasoline. I am changing the oil frequently in hopes of not killing the bearings.

The list of what I have checked and done so far:
new O2 sensor- changed mileage from 19 mpg to 24, not a huge change in performance, though it does feel a little sharper
checked throttle position sensor, intake air temperature sensor, map sensor- all in spec
did plug wires, cap and rotor, regapped plugs to .040-no significant change
tires are inflated correctly
Fuel pressure is in spec and does not leak down significantly with engine off, as tested several times, even overnight once. I believe this rules out a leaking injector, or a leak in any line between tank and fuel rail. I could be wrong about this. If anyone knows any other check relating to this, please chime in. Otherwise, fuel pressure seems to respond normally to changes in engine load and unplugging and restoring vacuum to the fuel pressure regulator, which I believe indicates correct function of the regulator.

Still unknown:
Is there an exhaust leak upstream of the first O2- I hope to answer this today
Is the ignition timing correct- I willcheck this today
Is there a leak in the fuel tank- I doubt it, but I intend to find out for sure
Would resetting the ECU be necessary/affect performance? If so, what is the best way to do so?

Any thoughts on this are appreciated. Would a leak in the fuel tank cause a CEL? Evap emissions or something? Is there an easy way to do a pressure test on the tank? Can anyone think of any other cause for rich running? Thanks in advance for any help you guys can give.

 

[AlaskaB16]

There is a fuse under the hood you can pull which will reset the ECU (Was on my Integra anyway), but I just unplug the negative battery terminal and let it sit for a bit when I need to reset. Is it possible you're just not burning all the fuel? Even if your plugs don't look bad, maybe replace then anyway? If your edges aren't totally square on them it will have negative effects on the burn. You sound like you're aware of that already though. Fuel problems are never easy to figure out as i've learned.

 

[Fusion158]

Maybe try a compression test on the cylinders to make sure that they are with in spec. Low compression would affect your MPG.

 

[upnsmoke]

Maybe try a compression test on the cylinders to make sure that they are with in spec. Low compression would affect your MPG.

That's a good thought Fusion. I would suspect something mechanical, except the engine seems to be running well. I have driven cars with engine issues, and they usually make themselves known, though not always. I know seat of the pants isn't exactly quantitative, and a compression check will be on the to do list. AlaskaB16, I may just replace the plugs to cover all my bases. My current ones are certainly less than perfect. I will check compression and leakdown at that point. I'm almost done getting the test pipe on, and have a wire harness finished for my wideband setup. I have to get some gaskets and hardware, but I will be able to get a read on AFR sometime today. Plugs and compression and leakdown test may get put off until tomorrow. Either way, I think the wideband will almost certainly tell me me something useful. Thanks for your thoughts guys. Please keep them coming.

 

[upnsmoke]

Well, fuel mixture is good enough under all conditions to convince me the car is running fine and I have just been chasing my tail here. I never saw fatter than 12.5:1 at wot and it was normally closer to 13:1. It was normally 14.5:1 to 15:1 at idle, and leaner than 14:1 while cruising at light throttle. That all seems about right to me. Compression and leakdown are in spec. Timing looked close enough at a glance, advance seems responsive. On a side note, the exhaust sounds nice without the cat. It's a stock replacement exhaust, so it's fairly quiet, but still has a little bark when you get on it that was not present with the cat in place. I changed the oil and gave it to my father to drive for the rest of the week. I guess the 24 mpg could have been partly due to the computer adapting to the new O2 sensor. We will check the mileage again at the end of the week. I expect at least high 20s, as that is what the car has gotten in the past. I'll get back with an update this weekend. Fusion and Alaska, thanks for your thoughts, and sorry for involving you in a wild goose chase.

 

[Fusion158]

No problem at all, glad to see you might be at the end of your search

 

[AlaskaB16]

Thanks and good luck!

 

[JohnS.]

I would replace the spark plugs and spark plug wires to cover all your bases. Are the injectors dirty? Fuel filter too maybe?

And running no cat will affect your gas mileage. Because your car is OBDII, there is a secondary O2 sensor after your cat, which I'm sure you are aware of. You probably smell gas because you're running richer, probably because you have no cat.

I would personally re-install a cat. I'm almost willing to bet your mpg will go back up.

 

[00siboy]

If your a/f ratio is where you say it is, and with all the other testing, idk what else you could do besides just keep an eye on the mpg the next couple weeks.

 

[upnsmoke]

I would replace the spark plugs and spark plug wires to cover all your bases. Are the injectors dirty? Fuel filter too maybe?

And running no cat will affect your gas mileage. Because your car is OBDII, there is a secondary O2 sensor after your cat, which I'm sure you are aware of. You probably smell gas because you're running richer, probably because you have no cat.

I would personally re-install a cat. I'm almost willing to bet your mpg will go back up.

The fuel filter is relatively new. I have not made any attempt to clean the injectors, though I do have the means to do so should I become convinced it is necessary. I think they are okay for now, as the fuel mixture would likely reflect any significant malfunction, at least under some condition. The change in mileage actually occurred with the stock cat installed. It looked okay to me upon removal, not clogged or anything. I have certainly seen worse on other vehicles that seemed to be running reasonably well. As far as I know, the second O2 is there for no other purpose than to measure the efficiency of the cat. I actually unplugged the second O2 when I put the wideband in, and used the bung in the test pipe. I have read of a couple dyno tests that showed very little difference in performance between open exhaust, high flow cats, and stock cats. I don't remember any details, but it was enough to convince me to leave cats in otherwise stock cars if they are not a source of an exhaust leak, and feel no need to change a tune simply on account of exchanging a cat for a test pipe or vice-versa. I can probably dig up at least one of the articles if anyone is interested. The exhaust doesn't reek of gasoline, as it shouldn't when the engine is running clean. It was the oil that smelled. After driving, while the engine is hot, when you remove the oil cap and put your nose to the opening in the valve cover there is a slight odor of gas. That should not be present on a correctly functioning engine with a normal amount of blowby. The problem I'm having now is that the gasoline smell lingers for a while, not necessarily to any ill effect on the engine, though it does take away one way to diagnose an excessive blowby condition should the mileage not improve.

 

[Trekk]

What is the operating temp of the car? If the car is running too cold that could be your problem.

It would be kind of funny if after all those test, it was just that

 

[upnsmoke]

What is the operating temp of the car? If the car is running too cold that could be your problem.

It would be kind of funny if after all those test, it was just that

I'd have to laugh. The coolant temp on the dash gauge appears to be in more or less the same place it always has been. The coolant temp comes up reasonably quickly, and all radiator hoses get hot to the touch, which indicates a functional thermostat. Is there a coolant temperature sensor other than the one for the dash gauge? If so, I suppose that could be something to look at. I hadn't thought of it.

On a separate note, I installed the second O2 sensor in the bung in the test pipe. I did not expect any difference in performance, nor have I been able to detect any, but I did expect a CEL. I have read of some using a spacer, usually made of one or two spark plug non foulers, to move the second O2 from the exhaust stream just far enough to give the O2 sensor a false reading which mimics that of a functioning cat. I also know that there is a product which is intended to produce a signal that mimics that produced by the second O2, supposedly making the computer think all is well, regardless of what is actually going on in the exhaust. I have never used one, but read that they are marginally functional. The funny thing is, with the second O2 installed straight in the bung on the test pipe, the CEL has remained off even as the car has been driven quite a bit over a couple days, and according to my OBDII scanner, has no pending codes that would make it fail an emissions test. This leads me to believe that although the CEL will undoubtedly come on eventually, one could likely pass an emissions test by clearing the codes within a couple days of going to the test. I'm not sure, but this could also be an indication that the second O2 is not functioning correctly.

At this point, I feel it necessary to say I discourage anyone from making any illegal modifications to any car, including those discussed in preceding posts. It is illegal to modify emissions systems with the intent of impairing their function or disabling them on any car that is driven on public highways. I make no admission of traversing public roads with illegally modified emissions systems on my car, nor of having intent to do so. I had to say it.
All that said, I got a chance to hear the car from outside of the car, and even with a stock replacement exhaust, it sounds awesome without the cat.

 

[upnsmoke]

Well, we monitored the fuel mileage over another week and came up with 18.6 mpg. Upon learning of this I took possession of the car once again for further diagnosis. I removed the oil cap with the engine running and was quite disturbed by the volume of gas coming from the valve cover. I'm going to put a Honda pcv in it tomorrow, but I'm feeling pessimistic about it. I will check compression and leakdown if the pcv doesn't turn out to be a cureall. I think we are lucky we haven't blown a seal driving the motor like this. I think we will be even more lucky if the compression and leakdown are in spec. Edit- I already checked the compression and leakdown on this one and it was fine. I have too many projects. Well, I sure hope it's the pcv. Otherwise, I think maybe I'm nuts.

 

[upnsmoke]

Well, I'm still going around with this thing. I retested the compression and leakdown, and they are in fact in a normal range. There seems to be nothing wrong with the engine. Also, while driving the car, I noticed that with a full tank the gas mileage appeared at first to be okay. Then I left the car for about an hour while I was in a store, and when I restarted the motor I noticed the fuel gauge had moved significantly. I pulled the tank yesterday, and found that the undercoat on the bottom of it had turned sticky like tar. The undercoat on the rest of the car was not like this, except along the driver side of the tank. It would appear that a significant amount of gasoline has leaked from somewhere and wet the undercoat where it is sticky. I can not find any evidence that gasoline has been on top of the tank. It does not appear to me, after a fairly thorough examination, that any high pressure or return lines have leaked.

The charcoal canister is stained and appears to have been wet, though I can't be sure it was gasoline. It could just be spilled oil. I cleaned it up and will check it again. Anyway, I can't find any obvious source of a fuel leak. The filler neck and all connected tubing appear in decent shape and do not seem to be leaking. I suspect that something is amiss with the evap system which is allowing gasoline to escape somewhere. There is a small rubber line going into the framerail which appears, judging by the pattern of the sticky undercoat, that it could be the source of the leak. I disconnected the hose from the framerail, and wiretied it in a position that will allow me to see any puddle of gasoline it makes if it leaks. I will keep an eye on it after I drive the car again.

I will be too busy to look at the car again for a few days, so for now the car is reassembled for my father to drive. I told him to keep a fire extinguisher handy, which is wise to do anyway. I have made a habit of it since I witnessed a friend's 90 Civic burn on account of a faulty wiring harness. We are certainly asking for trouble driving around leaking fuel. It has been a while since I worked on newer cars, and I don't remember much about OBDII evap systems. I will have to do some research about what can go wrong and how to test it. I did mange to plug all the lines on the tank and put some air pressure in it. It held pressure, which tells me the tank itself is not leaking. If anyone can tell me anything about what can go wrong with our evap systems without turning on the CEL, what could cause the leak I am seeing, or any related tests I can use to check operation of individual components in the system, I much appreciate if you could help me out. I feel like I have a lead as to what may be causing my issues here, but only that. I have no idea how to proceed in diagnosing this thing without knowing more about it's evap system. If no one can give me any direction here, I may try to track down a good shop manual for the car. Any suggestions on that? Sorry for making another long post. Hopefully I get this thing straight soon. At this point, I don't think there is much left to be an issue other than the evap.
On a separate note, I replaced the cat-back with a stock replacement system and can not believe how much quieter the car is. I had not realized how loud it had gotten, as it happened gradually. I almost miss the sound of the leaking exhaust. It really sounded good.

 

[00siboy]

Good to see you are at least finding a direction on this.

Your evap system doesn't hold enough fuel to leak all over the underbody. Your charcoal canister is what holds the fumes, but it doesn't hold liquid fuel that would be the source of a noticeable leak. Evap systems can cause poor fuel mileage if there is a leak, but 9/10 times you will get a code set, especially if its a leak big enough to cause terrible mpg. Unfortunately the only way I know to test an evap system for leaks is with a smoke machine, which the average won't have. If you still think its evap related then your best bet would be to have a smoke test performed at a shop.

 

[upnsmoke]

00siboy, I think I have certainly eliminated a few variables here. I worked for a Toyota dealer a few years back, and am familiar with the smoke machine. I pressurized all the lines myself when I had the fuel tank out and could not find any leak. Whatever is leaking must be significant to lose so much fuel. I will eventually have someone check it for me if I can not find anything wrong with it myself. I just can't help but think there must be a major issue somewhere.

I remember once, years ago, my father was working on a wrecked Mustang. He restored it to running/presumably drivable condition and drove it. All seemed fine with it until later as everything cooled, it dumped what appeared to be a couple gallons of fuel on the floor through a port on the tank that had become disconnected from an evap hose. It would seem that something would have to be terribly wrong to dump so much fuel, but nothing was ever found to be wrong with it other than the obvious evap issues, and once corrected, the car never leaked fuel again so far as we could tell. That tells me that, although the Mustang is a different car, there is a possibility of losing a significant amount of fuel through a severely dysfunctional evap system.

I'm thinking there is a chance we have a line blocked, valve hung open or not opening, or maybe even a bad ECU. I can only speculate though. If there is one thing I don't know much about on this car, it is the evap emissions system. I was wondering if something like this-link-manual would give me the details on exactly what parts are used where in the system, if not how to test their functionality. If I just knew which parts were which, I could research each one to figure out how it works and how to test it.

I intend to research this system enough to gain a reasonable understanding of it. While this may or may not be necessary to fix my father's car, I intend to buy a Civic of my own, possibly of the same generation, and as I tend to buy cheap cars that often turn out to be basket cases, I will want to be fairly familiar with the workings of the car so that I will be competent in working on my own junk. While I was always able to fix evap issues, or any others, with the Toyotas I worked on, including some cars as new as 08's (last year I worked for Toyota), I never took the time to research and thoroughly understand how they worked. I always just complained about the concept of simplicity being thrown out the window by every car manufacturer in the world. I still maintain that complaint, but the fact is that new cars, in all their computerized complexity, are not going away, so I might as well become proficient in repairing them.

 

[upnsmoke]

Looks like the link doesn't work on the Helm manual. Item # is 61S0307. I looked all over the underside and underhood of the car today while running and after it was shut off. I still have not found a leak or anything unusual anywhere. I'm going to buy the Helm manual to hopefully aid me in further research and diagnosis, unless someone has a better idea. My father has the car again and we are tracking the gas mileage as he drives it.

 

[00siboy]

You said it was moist on the underside of the gas tank right? Seems like you have been pretty thorough but have you checked the fuel line fittings on top of the tank to see if anything is leaking? I am starting to think that it is indeed a fuel leak causing the bad mpg because every evap system I have dealt with seems pretty sensitive to leaks and will most of the time set a code, unless its leaking after the car is shut off like you said. Sounds like you are going to have to just keep an eye on it and keep checking for new leaks like you have been doing.

 

[upnsmoke]

It definitely appears to me that the bottom of the fuel tank and the driver's side frame rail of the car next to the tank have been wet with gasoline. I checked the lines the best I could. I does not look to me like any are leaking. I ran the engine with the car up on jack stands while I looked and felt around the tank. The other thing is, I don't see any evidence of a leak occurring anywhere over the top of the tank. Everything looks dry between the top of the tank and underside of the car. If there was a leak on the high pressure side, I believe it would show in the form of leaking down fuel pressure when a gauge is connected with the engine off. Once I do sufficient research, I want to thoroughly check the function and integrity of the evap system.

Another thing I have not yet considered is that the car has some aftermarket suspension bits that allow it to take a turn harder than stock. It also has stiffer than stock shocks, so I feel I have to consider the possibility of sloshing fuel somewhere it would not normally be on a stock car. However, it seems likely that someone else capable of harder turns than us (autocrossers, someone with sticky tires, etc) or with stiffer suspension than us would have an even worse problem than we do, which does not seem to be the case.

I forgot to mention also that I found the top bolt missing from the evap canister under the hood. I do not know if the bolt was left out when I finished the car a year ago, in which case it would have been that way since, or was left loose and backed out. Either way, it appears that the evap canister has been allowed to repeatedly contact the subframe below it. Considering that Honda saw a need to mount the canister on rubber bushings, I would say repeatedly bouncing it off a metal subframe would pose a threat to its functionality. I'm not sure how this escaped my attention for so long, but I think it is something I need to investigate.