Poor MPG follow on

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
This is a follow up to some other threads here and on the other forum - from the Live Data Analysis, to the Gas Cap Venting, etc, etc. This thread is also over there but I'm looking at as wide a net as possible, because I'm stumped.

Bottom line: Poor mpg in 1996 Land Rover Discovery.

Initial analysis: Engine running lean causing ECU to overcompensate and dump fuel.

Analysis based on: High Long Term Fuel Trims

Mitigation efforts: New plugs and wires. New fuel pump and fuel filter, new front 02 sensors. Replaced MAF with known and tested used one. Replaced Charged Air Temp Sensor (shop stated it was showing -36F even though my OBD scanner shows it working correctly - maybe a stored value I can't see?). Smoke tested intake system - shop declared 0 leaks detected.

Recently - all plugs pulled. All look nice and chocolate. All plugs refit and torqued correctly. All wires checked for closure. Wires sprayed with water to detect arcing. None noticed.

This weekend completed an 800+ mile road trip:

Here are my findings (for purposes of the thread we will assume miles per hour was between 60 and 70 (generally 65), I drive like I have an egg on the gas and brake pedals. Outside of a few hill climbs RPMs stayed around 2200-2500. It only pegged 4000 rpm once, and if it did downshift it was at 3000 or so rpms.

All Gas: EXXON Supreme 93 octane. I started the trip with a full tank and a bottle of Techron fuel system cleaner.

Fill ups: Generally before the Orange light came on but less than 1/4 tank. I know I could get around 235 miles before my light came on so I was going that far each time.

1st Leg: 213.6 miles. 18 gallons. Put in another bottle of Techron. LTFT was at 3.1%

MPG = 11.86

(2 bottles of Techron through the system now)

2d Leg: 234.9 miles. 19.5 gallons. No additives. LTFT was still 3.1%

MPG = 12.04 (negligible difference could be result of elevation changes, no gains -- note: barely 2/10ths of a mile difference, but 1.5 gallons difference)

3rd Leg: 235 miles. 18.7 gallons. (roughly half of this leg was with LTFT of 3.1. About 40 miles were in town, but not really stop and go. LTFT dropped at one point to 2.3. This morning when I started back home LTFT was at 6.3)

MPG = 12.5 However, LTFT 3.1 for half a tank, then 6.3 for half a tank. (Still negligible mileage fluctuation from start of trip. Strange thing is LTFT was 3.1 for 10 hours of driving. Then suddenly switched to 6.3 for 10 hours of driving. Never fluctuated).

Here I put in a bottle of that Berryman Chemtool fuel system cleaner instead of Techron.

It gets interesting here. I had noticed that over the entire trip I could get 100 miles with the fuel showing 3/4 tank full. At about half a tank I'm around 150-160 miles. Empty (or darn near the light coming on) I'm at the 235 or so indicated above. I'm thinking it's just a gauge problem, but this time I stop and test my theory.

Leg 4. 106.1 miles. Stop at exactly 1/4 tank to fill up. It takes only 6.6 gallons. LTFT is still 6.3

MPG = 16.07 miles per gallon! If it were a faulty gauge I should have needed more fuel than 6.6 gallons. My truck is getting 16.07 miles per gallon if the tank is full.

LTFT stays at 6.3. At 3/4 tank again I note I am at 102 miles into the drive. I don't stop because I want to see if maybe overall mpg has improved, or if there really is something going on between 1/4 and 0 tank.

Leg 5. 213 miles. 15 gallons. (tank is now between 1/2 and 1/4 tank)

14.2 MPG.

If I fill up at 3/4 tank I'm pulling 16 miles per gallon. If I go down to between 1/2 and 1/4 tank I'm at 14.2. If I go until just about the time the light goes on I drop to 12 mpg.

So what would cause this? All the driving was the same. All highway miles. Same route there and back. Something is causing the engine to consume more fuel the emptier the tank gets. Note I have taken the gauge out of the question. I know some vehicles have slow gauges, but mine actually has better MPG at a full tank.

This morning I drove it to work. 50 degrees outside temperature. 30 second warm up idle. Closed Loop. LTFT 6.3 still!

STFT did not rise above -- numbers for the entire ten minute drive. I monitored it very closely. A few times it was at 0.0 but mostly it was at -2.3, -13, -14 even, and then -2 to -5 or so. LTFT never changed whatsoever.

So overall high positive LTFT, but during this morning's drive STFT was completely negative.

Once my LTFT went up to 6.3% I actually was getting better MPG.

Staying 3/4 tank and above and I add 4-5 mpg to my best MPG.
 

Tugela

Well-known member
May 21, 2007
4,754
561
Seattle
I like your empirical approach. Your experience reflects my observations that the rate of change in the dash fuel gauge fluctuates. On the highway I usually get >300 miles per tank, but like you I get more miles out of the first 1/4 tank than I do the third. This has been the case with all 3 Rovers I've owned so I thought it was one of those inherent characteristics of the brand, like the clunking transfer case.
 

WaltNYC

Well-known member
Mar 3, 2010
707
135
NYC
... the rate of change in the dash fuel gauge fluctuates.On the highway I usually get >300 miles per tank, but like you I get more miles out of the first 1/4 tank than I do the third.

Mine too. The tank float is a very simple device. The surface area of the top of the pool of gasoline in the tank is greatest when full due to the shape and 'posture' of the tank. I'm guessing the sensor for float arm angle has a linear progression and that makes the difference.

Basically the gas tank is somewhat 'funnel' shaped and the therefore the 'water line' of the top surface of the gasoline in the tank will move faster as the tank is emptied.

This still doesn't account for the 4th leg in the example though with a smaller fuel use and add, the accuracy of the calculation is reduced a bit. Never exactly sure the fill level was precisely the same.
 

Tugela

Well-known member
May 21, 2007
4,754
561
Seattle
I'm guessing the sensor for float arm angle has a linear progression and that makes the difference.

This is the concept I was trying to express: non-linear rate of fuel consumption relative to linear measurement by the sensor. My brain is low on fuel this morning.
 

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
This still doesn't account for the 4th leg in the example though with a smaller fuel use and add, the accuracy of the calculation is reduced a bit. Never exactly sure the fill level was precisely the same.

True. I tried to keep fill ups the same. Fill until handle releases. Squeeze once more until handle releases.

While there would be some difference, not enough to make a 4 mpg increase in fuel mileage.

I'm wondering if somehow I'm getting increased pressure in the tank the less fuel and that's somehow affecting overall fuel pressure? If tank pressure is lowering fuel pressure I'd see lower mpg the higher the tank pressure.

I'm really at a loss. I love 16 mpg but stopping every 100 miles to get it will suck.

Also doesn't explain how mpg increased when LTFT increased (doubled).
 

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
Drive to and from work. Only 2 miles each way but 15 minutes a pop. 30 mins total run time.

STFT stayed completely in negative numbers.

LTFT did not change at + 6.3
 

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
Another test: engine off, key on: 38psi fuel pressure.

Engine on: idle 30 psi
Engine on: 2500 rpms 25 psi.

Clamp fuel return line either running or idle: fuel pressure 100+ psi (gauge only goes to 100)

Vacuum detached while running: 36-38

Ten minutes sitting engine off: 28 psi

20 minutes engine off: 20 psi

1.5 hours: 0 psi
 
Last edited:

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
Anyone got any other ideas?

What about the fuel pressure? How long should it hold? Are my drop offs pointing to an injector? Doesn't drop immediately but the numbers are above.

Two more drives today. LTFT still at 6.3. Short is generally negative though did have a discrepancy in the two banks where one was pos and one neg.
 

bsa_m21

Well-known member
Sep 27, 2016
70
1
Vancouver Canada
Come on now - British fuel gauges have always been notoriously inaccurate. :rolleyes:

Regardless, the sending unit may be linear, but the shape and shape and angle of the tank bottom means that the volume of fuel is not a direct correlation to the elevation of the sending unit float. At a full tank, the volume of fuel represented for, say 1/2" of float drop, can be more than a 1/2" drop once the tank is 3/4 empty. Manufacturers try to compensate for that either in the sending unit or gauge, but to mixed results.

Your gas consumption, given your driving style is likely fairly consistent, but most likely the gauge isn't. :eek:


Given that you threw misfire faults, I'd check your Cat's before moving on to anything else.

Regards,

M.
'96 Disco 1
'52 S1 80
'80 TR7V8
 

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
Come on now - British fuel gauges have always been notoriously inaccurate. :rolleyes:

Regardless, the sending unit may be linear, but the shape and shape and angle of the tank bottom means that the volume of fuel is not a direct correlation to the elevation of the sending unit float. At a full tank, the volume of fuel represented for, say 1/2" of float drop, can be more than a 1/2" drop once the tank is 3/4 empty. Manufacturers try to compensate for that either in the sending unit or gauge, but to mixed results.

Your gas consumption, given your driving style is likely fairly consistent, but most likely the gauge isn't. :eek:


Given that you threw misfire faults, I'd check your Cat's before moving on to anything else.

Regards,

M.
'96 Disco 1
'52 S1 80
'80 TR7V8

True. I've had bad gauges before where it would drive forever and then suddenly drop, but the fuel consumption per gallon didn't change - the gauge just didn't register appropriate to usage.

This one definitely used less fuel (+ 4 mpg) at 3/4 to 4/4 and after 3/4 got -4 mpg overall. I plan on testing it again of course.

The cats only have 25k miles on them. Not saying they're bad but that would suck if so. How do you check cats?
 

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
All that really matters in your mileage per tank. If you're trying to relate MPG per quarter-tank, then, well, good luck.

Help me understand why.

If I put 6.6 gallons in an empty tank and get 12 mpg and go 79 miles
Or
Put 20 gallons in, but only use 6.6 and go 105.6 miles and get 16 mpg that's a big discrepeny and I want to replicate it and understand why.

And why I get better mpg at higher LTFT.

Hell if anyone knows a shop in the Pittsburgh area let me know, I'll take it there because this is frustrating and the only theory I've heard that was reasonable (air leak) was tested and busted.
 

luckyjoe

Well-known member
Oct 10, 2004
455
118
New Jersey USA
The fuel gauge is just a visual representation of your fuel supply. It is not a metered flow rate. So, going from a full tank to 3/4, what the gauge "shows" and what your motor drank are not the same, and it gets even fuzzier when you compare that to your 3/4 to 1/2 tank "values".

My VW Tdi will get 400 miles to the first 1/2 tank, but no where near that on the second 1/2 tank. When highway driving through a full tank my MPG's didn't change significantly, it's just that the fuel gauge is only a visual representation.
 

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
The fuel gauge is just a visual representation of your fuel supply. It is not a metered flow rate. So, going from a full tank to 3/4, what the gauge "shows" and what your motor drank are not the same, and it gets even fuzzier when you compare that to your 3/4 to 1/2 tank "values".

My VW Tdi will get 400 miles to the first 1/2 tank, but no where near that on the second 1/2 tank. When highway driving through a full tank my MPG's didn't change significantly, it's just that the fuel gauge is only a visual representation.

But have you measured the actual mpg you get on + 1/2 vs - 1/2 tank?

Take the gauge out of the equation: When full I get 16 miles to the gallon (using only 6 gallons). I'm taking the gauge out. If I had no gauge and stopped after 105 miles I used 6 gallons. If I only put in 6 gallons in an empty tank I get 12 mpg (or 70-something miles travelled). Add to that my extreme tank pressure when near empty (huge rush of air once I can vice grip the cap off).

More data:

1st bank has STFT in negative numbers. O2 1 and 2 fluctuate.

2d bank has STFT in positive numbers. 02 1 fluctuates. O2 2 stays at 1.005 to 1.010

I know the second 02 shouldn't affect fuel trim but seems something is there.
 
Last edited:

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
Based off some more research my 02 bank 1 Downstream 02 sensor should not fluctuate but stay static.

It definitely fluctuates up and down.

Bad cat it seems, but that doesn't explain bad mpg does it?
 

luckyjoe

Well-known member
Oct 10, 2004
455
118
New Jersey USA
But have you measured the actual mpg you get on 1/2 vs - 1/2 tank?

No, because the 1/2-Tanks indicated are not the same volume. The only way to know that each fill-up is identical, is to fill until I can see the diesel at the same point in the filler neck, which is a pain and takes too much time. And actually, why would I care? When I track MPG, I only calculate on a full-tank/fill-up data - especially long-haul/full-tank-consumption trips. For reference, this is the best MPG my 1995 RRC LWB has achieved (4-people, gear, long-haul driving through a complete tank):
attachment.php


When full I get 16 miles to the gallon (using only 6 gallons). I'm taking the gauge out. If I had no gauge and stopped after 105 miles I used 6 gallons. If I only put in 6 gallons in an empty tank I get 12 mpg (or 70-something miles traveled).

This is just too fuzzy for me - was it the same stretch of road, same conditions, same throttle activity, same payload? I suppose you could start with a dry tank, measured fill, and subsequent drain/measure of remaining tank contents - and get an exact MPG for each 1/2 tank. Yes, this is way-over-the-top and exactly why I only calculate on a full tank.

No ideas here on your tank pressure and fuel trim numbers...
 

AbnMike

Well-known member
Apr 6, 2016
1,218
117
Western Slope, CO
This is just too fuzzy for me - was it the same stretch of road, same conditions, same throttle activity, same payload? I suppose you could start with a dry tank, measured fill, and subsequent drain/measure of remaining tank contents - and get an exact MPG for each 1/2 tank. Yes, this is way-over-the-top and exactly why I only calculate on a full tank.

Same road (going opposite direction), same throttle activity, temperature, payload etc.

If there wasn't such a huge discrepancy I wouldn't be chasing this rabbit. But an extra 4 mpg is huge for me.

Using 19 gallons (tank) I get 12 mpg overall.
Using only the top 6 of those 19 gallons I showed 16 mpg.

Anomaly? Maybe. I'm doing further testing of course. But when I think of stuck fuel cap, huge rushing volume of air when removing the cap, etc I'm chasing rabbits...

12 mpg fucking blows and I can't figure out the cause. There's no explanation I haven't tried and by lack of responses I'm guessing no one else knows either. It's frustrating that I'm driving the worst mpg Discovery out there but is as tuned up as I can make it. Heads, cam, exhaust, etc etc all have 25k miles...