Mystery of Bad Gas Mileage

[Lawsonmh15]

Well, I've put 8,000 Miles on my "new" 99' Legacy SUS in the last couple of months. I know that the best mileage to hope for is 28/21. Generally I stay in the 25-27 Mpg range. However, In the last week I drove to the land of Corn and cheap gas... Nebraska. Filled up on 87 at 30 cents cheaper than anywhere in CO but was treated to 20MPG pure highway. Having only had the car a couple months, I am EXTREMELY picky about it's performance and noises. I spent the better part of two nights tuning my throttle cable and resetting the ECU. My dad is the same way with his motorcycles and vehicles. After screwing with my head with suggestions of pump, vac, o2 sens, probs; he was nice enough to point out that ALL GAS with 10% ETHANOL (corn based) will net a MUCH lower mileage. At the next stop, Filled with 85, no Ethanol, and back to an appropriate range.

 

The moral...

 

If you are freaking out over MPG and not paying attention at the pump, perhaps it is YOU to blame, not your Subie;)

[friendly_jacek]

Well, I've put 8,000 Miles on my "new" 99' Legacy SUS in the last couple of months. I know that the best mileage to hope for is 28/21. Generally I stay in the 25-27 Mpg range. However, In the last week I drove to the land of Corn and cheap gas... Nebraska. Filled up on 87 at 30 cents cheaper than anywhere in CO but was treated to 20MPG pure highway. Having only had the car a couple months, I am EXTREMELY picky about it's performance and noises. I spent the better part of two nights tuning my throttle cable and resetting the ECU. My dad is the same way with his motorcycles and vehicles. After screwing with my head with suggestions of pump, vac, o2 sens, probs; he was nice enough to point out that ALL GAS with 10% ETHANOL (corn based) will net a MUCH lower mileage. At the next stop, Filled with 85, no Ethanol, and back to an appropriate range.

 

The moral...

 

If you are freaking out over MPG and not paying attention at the pump, perhaps it is YOU to blame, not your Subie;)

 

Thanks for pointing out the common knowledge fact.

Ethanol has roughly half of the energy per volume unit compared to gasoline. So, with 10% ethanol fuel , you can expect 5% loss of MPG, and power, maybe little bit more due to the need to dowshifting. The change you observed was more of underfiling with subsequent overfilling or perhaps cheating at the pump.

 

I use 87 gasoline with 0% alcohol and still experience 16-20 MPG city and 22-24 MPG hwy.

 

Any suggestions for me?

[Lawsonmh15]

Thanks for pointing out the common knowledge fact.

Ethanol has roughly half of the energy per volume unit compared to gasoline. So, with 10% ethanol fuel , you can expect 5% loss of MPG, and power, maybe little bit more due to the need to dowshifting. The change you observed was more of underfiling with subsequent overfilling or perhaps cheating at the pump.

 

I use 87 gasoline with 0% alcohol and still experience 16-20 MPG city and 22-24 MPG hwy.

 

Any suggestions for me?

 

Possibly cleaning your MAF. I have had great success with the BG&K 402 "Brake" cleaner. It claims to be O2 sensor safe, so I just removed the MAF and sprayed a light stream over the screen and resistors on the MAF. Weird hesitation decreased dramatically. The rest was fixed after removing all slack AND wiggle room from my throttle cable. Fuel filter would be good to look at replacing as well. I try to look at what are most likely the oldest items on my car. i.e. Spark Plugs, Plug wires, Air Filter, Tranny Filter, Front/rear diff oil and then some normal things like tire pressure.

 

Also, if you decide to clean your MAF and are taking off the first air intake, make sure to look inside the second resonator air intake to see if there is oil present. If so, you will need to replace your PCV Valve $4 clean the hose and air intake. Then you will want to run a can of 44k through a quarter tank before your next oil change. Without knowing what engine you have, I can't say for certain what you should be at, but if it's a 2.5L or less, you are hurtin. Get codes pulled at a checker or napa and if it comes back PO130, you may want to consider replacing your O2 sensor. I just picked up some 60 degree offset metric wrenches for doing my AC Retention bracket, but haven't tried to remove the O2 yet. Would like to see if a difference could be had by cleaning it with the BG 402.

 

Any hesitation?

Smelling Rich?

Here a miss? (Check plug wires. Have had mine shoot off in the past.)

Rough Idle?

Changed parts without resetting ECU?

[friendly_jacek]

Thanks for the comprehensive listing.

 

1. I have MAP sensor and not MAF (MY 2000+).

2. Spark plugs, all filters, all fluids replaced at 30000, will replace again at 60000 soon.

3. Spark wires OEM but resistance checked and OK.

4. Tires inflated to 40-44 PSI (depending on load).

5. PCV checked and works OK.

5. Injector cleaners run a couple of times (redline and techron).

6. No OBD2 codes, but approx -10% long term fuel trim.

7. O2 sensors confirmed to work OK (replaced at 20000 due to a recall anyway).

8. Did I mention that all fluids are synthetics?

9. It does smell rich on cold starts, I thought fuel pressure regulator was defective, replaced, no change.

10. ECU resetted a few times.

11. BTW, it is NA 2.5L, phase 2 (SOHC).

 

I had the car since new, and the poor MPG was from day 1, dealer could newer find a problem, except for defective TPS, that did not affect the MPG after all, despite my initial impression otherwise. I run a couple of used oil analysis, one result showed fuel in oil (4%), but that was after idling. Wear metals (especially Chromium) were high normal, otherwise nothing obvious.

[Lawsonmh15]

Thanks for the comprehensive listing.

 

1. I have MAP sensor and not MAF (MY 2000+).

2. Spark plugs, all filters, all fluids replaced at 30000, will replace again at 60000 soon.

3. Spark wires OEM but resistance checked and OK.

4. Tires inflated to 40-44 PSI (depending on load).

5. PCV checked and works OK.

5. Injector cleaners run a couple of times (redline and techron).

6. No OBD2 codes, but approx -10% long term fuel trim.

7. O2 sensors confirmed to work OK (replaced at 20000 due to a recall anyway).

8. Did I mention that all fluids are synthetics?

9. It does smell rich on cold starts, I thought fuel pressure regulator was defective, replaced, no change.

10. ECU resetted a few times.

11. BTW, it is NA 2.5L, phase 2 (SOHC).

 

I had the car since new, and the poor MPG was from day 1, dealer could newer find a problem, except for defective TPS, that did not affect the MPG after all, despite my initial impression otherwise. I run a couple of used oil analysis, one result showed fuel in oil (4%), but that was after idling. Wear metals (especially Chromium) were high normal, otherwise nothing obvious.

 

Hmmm... You sound like you've tried quite a few things. I would talk to the Subie genie in the magic jellybean field and request better gas mileage. Strange that the mileage has been that way from day one. You stated that you had gas in your oil? My reasoning says that it's not being fully detonated. Have you checked to see if you have oil in the second resonator air intake? If not, you should. You could be gumming up your injectors and farming gremlins. I run semi-synthetic and have never experienced anything but a smoother feel and improved gas mileage, so I have to imagine that a full syn wouldn't cause this issue. I am not a mechanic by any means, but have an aptitude for researching, following diagrams, and figuring ************ out. It sounds like you may exhibit the same traits. I don't know that I would trust many others opinions or inclinations than my own... Neither should you. It sounds like you'll figure it out. Good luck.

[idosubaru]

i'd check your brakes all the way around. make sure the slides on the caliper are well greased and the brake pads are wearing flat and even.

 

my 97 OBS was getting poor mileage and i replaced all tune up parts, timing belts, etc with no change. so then i attacked the pads (eventhough they didn't need it) and the mileage went up a few mpg. i really went into the brake pads only because i had never inspected them since i bought the car. didn't really think about mileage but noticed it increased after the brake job.

[1 Lucky Texan]

I think folks have experienced the engine temp sensor (not the temp gauge sensor) being bad and keeping the ECU fooled into a 'choke' condition.

 

Carl

[friendly_jacek]

Replaced all brake pads at 46000 miles, no unusual wear.

The engine temp as checked via OBD2 is correct.

I covered all the basis and still clueless about it.

BTW, wheels are perfectly aligned (took several tries though).

I suspect that the -10% long term fuel trim has the answer.

I measured fuel pressure and it appeared 10PSI too high, but didn't change when I replaced fuel regulator.

[Lawsonmh15]

Replaced all brake pads at 46000 miles, no unusual wear.

The engine temp as checked via OBD2 is correct.

I covered all the basis and still clueless about it.

BTW, wheels are perfectly aligned (took several tries though).

I suspect that the -10% long term fuel trim has the answer.

I measured fuel pressure and it appeared 10PSI too high, but didn't change when I replaced fuel regulator.

 

You said yourself that it smells too rich. Is this just on startup, or does it lug down until about 3,000 then take off like a rocket? What plugs are you running. If you chose, or were sold , plugs that were a "step colder" that may be your biggest problem. I was going to purchase some NGK Iridiums but couldn't afford to wait for them, and bought the Bosch platinum +4's(No probs whatsoever). But I have heard horror stories of people purchasing colder track plugs expecting high RPM.

 

The other thing I was thinking of, and you may have already alluded to this, but what is your idle after warm up and in traffic. On start I am at 1500-1600, then rev to 2500 and drops to 1000-1200. Put in gear I go to 800 (Which I don't like because of the sound when putting into gear) but after cruising for a few, I stay at about 350-400 W/O AC on in gear at a full stop. I have heard people say that normal idle is 600-800 warmed up, but I don't believe that's right for the 2.5. I experience NO shudder whatsoever.

 

So, If you are always idling high, I would imagine that the ECU is not recognizing that the engine is warm.

 

The other thing is that you might want to check the tension of your throttle cable. I fiddled with mine until I got smart and let run after warmed up and adjusted until I heard the idle raise, then backed off until drop to normal.

 

I know people are offering alot of opinions about unrelated items, but it's been my experience that you need to look at direct connections.

 

My final offer is power. Proper voltage out of coil pack? Proper current on alternator? I know it sounds nominal, but you should check the grounds from battery, to chassis. If you notice an increase in headlamp brightness under increasing RPM, you may have dirty/bad grounds. Also, if you have alot of thumpity stereo gear being run, you may not be providing enough current for all other electronics to properly function. This is also true of a poor ground. From the factory, I noticed that the ground points on the chassis were painted?!!! That's useless. I spent about 20 minutes scraping paint around ground points in order to get a TRUE ground.

 

It sounds like you've tried ALOT of things, and I understand your hesitation to want to go out and continue trying other things, but everything I have mentioned is either free, or under $50 to try. I would advise you try the quickest and easiest things first, but If your idle is constantly high and your throttle cable doesn't make any difference when adjusted, or has obvious slack, then I would be inclined to believe that it is the sensor that tells your ECU that your engine is warm and to cut the idle.

 

Also, is your throttle sticky? If so you may just need to hit the coils, pivot points, and cable itself with some lithium grease.

 

Lastly, what vehicle is this exactly, and where do you live (elevation).

 

Good luck my man.

[Olnick]

Excellent suggestions, Lawsonmh15. Well thought-out, logical and basic (good old "KISS")!

 

Your observation about the ECU not sensing engine temperature properly is right on. When I replaced my engine coolant temp sensor it made a fantastic difference in smoothness and idle speed. Thankfully I got a code that told me to change it--I probably never would have figured it out on my own!

 

Aloha.

[Lawsonmh15]

Excellent suggestions, Lawsonmh15. Well thought-out, logical and basic (good old "KISS")!

 

Your observation about the ECU not sensing engine temperature properly is right on. When I replaced my engine coolant temp sensor it made a fantastic difference in smoothness and idle speed. Thankfully I got a code that told me to change it--I probably never would have figured it out on my own!

 

Aloha.

 

Thanks man. I hope to move to Kona in a couple years, maybe we can hook up.

 

Just another point I've learned. Olnick got a code. I've heard of alot of people NOT getting codes, or false positives. I've heard of people running with the o2 sensor disconnected (rectifying their rough idle/ bad MPG and resulting in MUCH better performance) but never got a CEL even though the O2 WAS bad and was disconnected. Computers are great for A/F calculations and many other binary functions, but nonetheless, it is NOT infallable.

[friendly_jacek]

Thanks Lawsonmh15, I will try these things, nothing to loose. I have a new set of NGK double platinums on hand, just waiting for the 60000 mark. I have heard good things aboud ground mods. Will do for it for fun.