FAQ: Improving Bad Gas Mileage

AA wrote:

It does happen to a small degree with a cooler t-stat, but it is a negligible difference. Spark plug heat ranges do not make a difference with regard to economy.

The theory is, the motor is engineered to optimally operate at a specific temperature. For us, the temperature is ~200ºF. At that temperature, all components in the engine and transmission should operate at peak efficiency. When you lower (or raise) engine temps, you leave the range of peak operating efficiency, resulting in slightly decreased economy.

A cooler t-stat decreases coolant temps, which forces the engine to operate at a lower temperature. A 180º t-stat is only down a slight bit, and I haven't noticed a difference in MPG. Colder than that will cause the engine to run at a temperature close to warm-up cycle, causing poor efficiency and much more wear to the engine's internal parts.

Cooler range plugs affect combustion temps, but have no real effect on the overall operating temperature of the engine, so should not impact MPG.

The reasoning behind using cooler plugs and t-stat is to lower KR (knock) levels, which makes the engine better for safely producing power and torque under boost, but this does not make the engine more efficient for fuel economy, which really applies only when you are not running much boost (lower RPM, less load).

Cliff's notes: Conditions for peak power efficiency ≠ conditions for peak fuel efficiency.

Hope this helps.

For those of you reading this who are thinking about going to university to study engineering -- this is a really cool problem. Theoretically it should be that if you're generating peak power it's because you are using all the fuel that you're inputting into the "system." In engineering you'll learn why this is not always the case and perhaps you'll come up with some workable ways to address that problem. Give you a hint: has to do with friction, moving parts, and how much of the fuel input ends up as heat instead of motion...

It's a 8very* interesting problem, you could make a career out of working towards solving it via fundamental changes in design, materials, and systems. For instance, if you run an internal combustion engine at peak efficiency to generate electricity, you can modulate traction output with less waste by using that electricity to power motors for movement. When not moving you are running engines to store power in batteries or capacitors, not wasting power. When you need a lot of power you can put more juice into the electric motors and still be on a reasonable efficiency curve. Without going into more details, this is why diesel-electric designs for locomotives have dominated freight railroading for some time now. Does this imply something similar for cars?

Think about it.

Albertj

Good points, Albert. From what I know, our engine likes a certain air/fuel ratio to make max torque, and when combined with RPM, max power. This ratio is right around 11.5:1. This is not the same as the air/fuel ratio used for max efficiency, which is 14.7:1 (less fuel for a given amount of air).

My research indicates the 14.7:1 ratio was chosen because it allows the catalytic converter to operate at optimum efficiency. Before cat converters were employed, some companies were experimenting with "lean burn" gas engines. I think Honda actually manufactured them in the '70s. These lean burn engines could run at around 20:1 air/fuel, resulting in exceptional MPG. My understanding is they did not build good torque at this ratio, so under throttle, the mix was enriched for better power.

Today the lean burn trend has returned with the more common use of direct injection technology. Hyundai is leading the charge in this area, among others. A special cat converter system allows the air/fuel ratio to run very lean when there is a potential for economy driving. In theory ratios as high as 60:1 are possible. I wonder if our 3800 engine, with cat removed, could be adapted for lean burn at cruise? I've often thought about doing it, as it only requires some PCM tuning, but I don't know enough about the other dangers that may exist. Thoughts?

AA wrote:

Good points, Albert. From what I know, our engine likes a certain air/fuel ratio to make max torque, and when combined with RPM, max power. This ratio is right around 11.5:1. This is not the same as the air/fuel ratio used for max efficiency, which is 14.7:1 (less fuel for a given amount of air).

My research indicates the 14.7:1 ratio was chosen because it allows the catalytic converter to operate at optimum efficiency. Before cat converters were employed, some companies were experimenting with "lean burn" gas engines. I think Honda actually manufactured them in the '70s. These lean burn engines could run at around 20:1 air/fuel, resulting in exceptional MPG. My understanding is they did not build good torque at this ratio, so under throttle, the mix was enriched for better power.

Today the lean burn trend has returned with the more common use of direct injection technology. Hyundai is leading the charge in this area, among others. A special cat converter system allows the air/fuel ratio to run very lean when there is a potential for economy driving. In theory ratios as high as 60:1 are possible. I wonder if our 3800 engine, with cat removed, could be adapted for lean burn at cruise? I've often thought about doing it, as it only requires some PCM tuning, but I don't know enough about the other dangers that may exist. Thoughts?

No further thoughts, my understanding is that you need to make changes to crank and cam as well for correct lean-burn but I don't know any more authoritative info about it.

Albertj

AA wrote:

Good points, Albert. From what I know, our engine likes a certain air/fuel ratio to make max torque, and when combined with RPM, max power. This ratio is right around 11.5:1. This is not the same as the air/fuel ratio used for max efficiency, which is 14.7:1 (less fuel for a given amount of air).

My research indicates the 14.7:1 ratio was chosen because it allows the catalytic converter to operate at optimum efficiency. Before cat converters were employed, some companies were experimenting with "lean burn" gas engines. I think Honda actually manufactured them in the '70s. These lean burn engines could run at around 20:1 air/fuel, resulting in exceptional MPG. My understanding is they did not build good torque at this ratio, so under throttle, the mix was enriched for better power.

Today the lean burn trend has returned with the more common use of direct injection technology. Hyundai is leading the charge in this area, among others. A special cat converter system allows the air/fuel ratio to run very lean when there is a potential for economy driving. In theory ratios as high as 60:1 are possible. I wonder if our 3800 engine, with cat removed, could be adapted for lean burn at cruise? I've often thought about doing it, as it only requires some PCM tuning, but I don't know enough about the other dangers that may exist. Thoughts?

Im not going to be the Guinna pig on this one. Having AFR at say 50:1 would mean the injectors are barely spitting out fuel. I could see even stock injectors having a really hard time atomizing the fuel at that level for fueling. Plus the dangers of KR spikes and a lack of power when you hit hills. IMO adjusting the cruse AFR that extreme would be similar to making you timing 80° at cruse. Jumping on the throttle would command the car to start pulling timing and at least with my 96, the pcm doesnt pull it anywhere near fast enough to go from 80°-16° without having problems (KR).
Could it be done on a 3800? I dont see why not. but i dont think it could be done and still keep power.

I was thinking more along the lines of trying 16:1 - and only in situations where quick changes in A/F probably won't happen, like for example when the timing advances up to 38-40º. It would just be a way to sip a little less when the PCM tells the engine to sip. If it worked, maybe try 18:1 and watch for knock.

The thing I'm worried about is what happens on a long term cruise at lean ratios. Will internal parts (valves, pistons, manifolds) get hotter in operation?

possible plug fouling too with the 104's?

Not sure. If I cleaned them off with some WOT once in a while, might be all right. I do think your right they'd foul faster than usual.

Howdy

I've done all the maintenance suggested here that I can find to improve my mileage which has fallen from about 26.5 mpg (highway - 60 mile per day round-trip ) on average to something around 20 mpg. I guess I don't maintain the car as well as I tell myself I do.

These are the things I've done in order and about the improvement in mileage I obtained:

Maintenance

Baseline 20.xx mpg
Cleaned air filter, inflate tires, Seafoamed gas tank and crankcase
New NGK platinum plugs 21.88
Replaced Delco fuel filter with Napa Gold, added 1 pt tranny fluid 22.55
Tested Bosch O2 sensor w/torch, cleaned MAF sensor with contact cleaner 22.98
Confirmation Calculation: 102.3 mi / 4.6 gal 22.73
Replaced O2 Sensor with OEM A/C Delco, 1 rear tire @16 psi 65/2.58 25.19 (dubious number)
Seafoamed intake- 136 miles/ 6.15 gal 22.11

So, I started at about 20 mpg and am still at about 22 mpg. Tires are inflated to 32 psi.

I'm not sure what else I might do to get the mileage back up. I've gotten as much as 28.5 mpg on long trips before.

Can anyone provide some advice on what to do next ? I'd appreciate it. The car runs great it just doesn't get the mileage I think it should.

No SES light. I might have gotten a little high-temp antiseize on the O2 sensor when I installed it. Is it possible it failed ? I don't have a scanner yet and wouldn't really know what to look for anyway. Is it time to bite the bullet and get a scanner ?

Thanks in advance.

...you have a '96 supercharged Riv...

I have some 'starter' questions.

1) which O2 sensor did you replace, the upstream one or the downstream one? [the downstream one is the one under the car and essentially all it does is tell the PCM that the cat works]

2) what brand and type (synthetic or not) of oil do you use? [synthetics are usually slicker]

3) is your 'service engine soon' light on? [indicates other issues]

4) what RPMs is your car running when cruising at 55 mph? 65 mph? [indicates transmission issue, for instance the TCC may not be working right]

5) Have you ever removed and cleaned the throttle body? [removing it is a pain but on this car it's really the only way to clean it right and as well to clean the EGR passage in the SC]

6) have you made any changes to the programming of the PCM?

Albertj

Thanks Albertj

1) Replaced upstream O2 sensor (in downpipe ?)

2) I've always used Castrol GTX 10W-30 conventional oil.

3) The SES light is not on.

4) It's typically in overdrive at those speeds - I'm guessing 1500 rpm at 55 and about 2000 rpm at 65 mph.

5) No, I've never removed and cleaned the throttle body. When I Seafoamed the top end of the engine it didnt smoke as much as I was expecting it to. So, it wasn't that dirty ? I don't know.

6) The PCM hasn't been reprogrammed.

The car seems to run as well as it ever has - good throttle response and power, transmission seems to shift where it should - that improved with the new fuel filter to my surprise.

One thing I forgot to mention - I drove the car for about three weeks without the S/C belt on until I had time to change it. Could that have contributed to intake fouling of some kind ?

Thanks

Quick reply -

Consider replacing the other O2 sensor, maybe get a Denso from RockAuto they work *very* well and are less than $35.

Consider removing and cleaning the throttle body. You'll be dismayed at how foul it is on the other side o' the throttle plate.

If you put a good synth in (Mobil 1, Castrol, etc) you will hear the difference even at idle. Make a recording before and after...

2000 RPM at 65 MPH sounds high. I wonder if the lockup clutch is slipping, or not engaging for some reason.

Use cruise control as much as you can on the highway if only because (according to the service manual) the PCM succeeds better at optimizing fuel trim with it on than without it.

Remember that the gas you use has an impact and mileage goes down in winter due to winter gas. Ethanol blends get worse mileage too.

What else can you tell us?

I noticed that some of those reading seem to be from very short trips which is going to affect the accuracy of the readings. Unless you're running the exact same route at the same speed every time its an apples to oranges comparision. Consider for instance when refueling, one day after the pump stops the guy just keeps clicking away to round up to the next dollar and the next day he doesn't. If you use a higher mileage test that won't effect outcomes as much. Sometimes its the little things can kill mpg. Has there been more traffic recently? Was a new light added in your route? Windows up or down? Was the defrost on at any point? Defrost engages the a/c which lowers mileage. Albertj is right more info is needed to help sovle this.

with Derek. Your MPG reports are right in sweet spot for a Riviera, but they're all acceptable. Driving habits, weather conditions, and route make a much bigger difference than mods or maintenance, even more than windows down or AC on (both of which really don't matter much with our car). Tire pressure is another one that makes a difference.

Here are some approx MPG measurements I typically see over the course of a year's time, driving the same car:

~ 16 MPG or less: racing
~ 17 MPG: lots of city driving
~ 18 MPG: normal route + racing or towing trailer
~ 19 MPG: normal route in snow/ice conditions
~ 20 MPG: normal route in avg winter conditions
~ 21 MPG: normal route avg year-round
~ 22 MPG: normal route summer conditions.
~ 23 MPG: normal route + hyper-miling strategies
~ 24 MPG: normal route + long distance trip
~ 25 MPG: normal route + extra long trip
~ 26 MPG: very high speed long trip (80 MPH)
~ 27 MPG: high speed long trip (75 MPH)
~ 28 MPG: normal long trip (70 MPH)
~ 30 MPG: normal long trip (65 MPH)
~ 32 MPG: low speed long trip (60 MPH)
~ 35 MPG: lower speed long trip (55 MPH)
~ 37 MPG: low speed long trip + hyper-miling (-50 MPH)
~ 38+ MPG possible with tuning and extreme hyper-miling practice

Seems like a wide range, but each is entirely repeatable based on route, conditions, and driving techniques. The car, engine, and tuning are exactly the same for every one of these measurements. Most dips in economy can be explained by a change in driving routine. A consistently lower number could mean something is wrong with the car, but I've found that usually isn't the case. If 38 MPG seems hard to believe, I've repeated it 3 times on a 120 mile route. What you need to do to get that kind of efficiency is unbelievable, but I did it to show it could be done. Can I get that everyday? Nope, nor would I ever try. All the conditions need to be absolutely perfect, and that's just not realistic in Ohio.

For more on techniques to improve economy, read this thread:
https://rivperformance.editboard.com/t3481-faq-hypermiling-modifications-techniques

Thanks for the replies all.

I realize the short mileage measurements aren't precise - they're merely a way to get me a ballpark estimate quickly instead of burning a full tank before I realize I need to do something else. Driving this morning the rpm was 1800 at 65 mph and about 1500 at 55 mph.

The fuel around here as ethanol in it - there are few stations without it. But that's the way it's been for some time - at least a year or more. But maybe they've upped the percentage - that could be.

The route and usage of the car has been pretty uniform since I got it in '05 with 92k miles on the odometer. 30 miles each way to work and back 5 days a week on the freeway except for about 3 of those thirty miles which are on feeder routes. Initially on that route I was getting in the range 25.5 - 27.5 mpg. If it went below about 25 mpg it meant the air filter needed cleaning and/or the tire pressure was getting low. Typically I don't use the cruise control on that commute.

Regardless, I think I will try some synthetic oil - probably should have been using it all along. And I t hink I'll order a new Denso O2 sensor. I replaced the original 02 sensor at about 104k miles with the Bosch unit - I didn't need to as there was nothing wrong with the old one. The big mistake is I threw away a known good one with which to make comparisons Before I replaced the Bosch with the A/C Delco, I used the propane torch test on the Bosch - output was around 925 mV consistently over two minutes and would go below 450 mV almost instantly when the torch was withdrawn. My one doubt over the Bosch was that the voltage output would "flicker" below 450 mV for a fraction of a second when I was holding the torch as steadily as possible - but I thought this could be attributed to even minor flame disturbance. But I decided to change out the sensor to the Delco anyway because, hey, then I'd get my "spare" 02 sensor back, right ? And, it was possible that that "flicker" was not me moving the torch but a sensor that was going bad. I still don't know. The Delco seemed to help for one short (inprecise) measurement but now does not seem to have made any difference.

AA, thanks for the link to the mileage thread - I have seen and read that one.

Does cleaning the TB and the EGR passage help with improving mileage ? I was surprised that a clean fuel filter helps with mpg - but after thinking about it I shouldn't have been. My old one was quite restricted.

Eric, it will be interesting to see if you see any improvements using different oil and O2 sensor. I have nearly the exact same commute as you, 30 miles each way 5 days per week, plus a little town driving on the weekends. It's 70% freeway by my estimate. I use an AC Delco O2 sensor installed in '05, and conventional 10W-30 oil. With the listed mods, I've logged an average 21.5 MPG combined over a 3 year period. Any time MPG was higher or lower, it was for one of the reasons listed above.

This might surprise a lot of folks, but I have never cleaned my TB or EGR for the life of the car (226k miles). I've never done a top end clean with Sea Foam or similar product. And even with a fender well mounted air filter, which gets pretty dirty, I don't notice a whole lot of difference in economy between cleanings, which I try to fit in every 6 months.

I've heard a lot of talk about cleaning the tb but thought 'eh its only got 72k on it, how dirty could it be?' Boy was I wrong! My car was always well maintained and super was always used but that tb was DIRTY! There was a thick layer of soot from the throttle plate back, so much so that in places the exterior of the motor was actually cleaner. I'd post pics for educational purposes but I don't know how. Just goes to show how much carbon the egr let's thru. You'd think there would be a replaceable filter on it or something. Plus after 'only' 72k the tb gasket was shot, it appeared to have broken into two pieces, which we all know causes it to hunt for idle speed. If you have the time and the skill it could only make things better to clean the tb and replace the gasket.

I agree, I should do it. Only reason I haven't is because the car runs really strong as is, and idle is smooth, so maybe I got lucky. I also don't know if being dirty affects performance as much as some think. I know it looks bad, but what do the numbers say? No one really knows. I guess I let mine go to see if things would deteriorate, and they really haven't, or so it seems. Having said that, when I get to cleaning the thing, I'm sure it will seem like a boost in power, and maybe economy. Putting on my list...

Derek wrote:

I've heard a lot of talk about cleaning the tb but thought 'eh its only got 72k on it, how dirty could it be?' Boy was I wrong! .

So Derek,

Did you notice a performance improvement after cleaning the TB ?

Unfortunately I will never know. I was in an accident and am in the process of parting out my rivi so I didn't find out till it was too late. I would guess no tho because on a stock car I don't believe the tb to be the most restrictive part. I would bet that the stock tb will outflow the stock air cleaner assembly, so untill you open up the intake it will remain the obstuction and not the tb. On a modded car the result may be more noticeable. Throttle response may be marginally improved but I would doubt you'll pick up much lost hp. If I recall correctly you're at 173k, does your cars idle remain steady or fluctuate?

Sorry about the deer attack. I wish they wouldn't do that.

The idle is pretty smooth and steady now with the new plugs and fuel filter.

Thanks for the kind words. As far as I'm concerned now all deer are good for is venison lol. Do you by chance have a scanner? The best way to check your idle is to plug in and watch for fluctuations. If it really is running steady I wouldn't bother pulling the tb just to clean it. But if not then changing the tb gasket and cleaning go hand in hand. Be careful of the maf! Only use the proper cleaner on it or you will damage it.

I was going to get an HP Tuners years ago, then wait, maybe DHP would be better ? Then, I can't decide. Maybe the freeware version with diagnostic cable for starters until I get around to doing the research. I still don't know how the support for the ' 96 stacks up amongst the vendors. I should probably get serious and find out. I've read the threads on the subject but they're anything but clear-cut. At least to me.

Don't get confused between a scanner and a tuner. A scanner will read what the pcm is seeing, a tuner can read and make changes to the pcm. You listed tuners which, unless you plan on modding your car, are more tool than you really need. For basic readings and diagnostics a scanner like actron is more than enough and can be invaluable when trying to track down problems. Scanners can be had for $200-300 or so but tuners will cost about $500+. And one of the tuners (I think hp) does not support the 96.

Yes, I do tend to mix up scanners and tuners. My thinking was "why spend $200-$300 dollars for a scanner when you could spend $500 and get something that would do everything you'd ever want ? " It's just money, right ? I have been considering some mods. An intercooled 300 hp Riviera would be pretty sweet. But maybe just a pipedream too. It's not a priority.

Reporting back as ordered, Sirs:

I changed out the AC Delco front O2 sensor to a Denso and achieved this mileage:

365/15.7 = 23.25 mpg August 13th - Normal workweek commute 30 miles highway one way.

Then I changed out the Castrol GTX 10W-30 to Mobile 1 synthetic 10W-30

433.8/18.3 = 23.77 mpg September 29th - Normal workweek commute 30 miles highway one way.

So no appreciable difference was obtained. Has anyone who cleaned / changed out injectors noticed a mileage improvement ? I would think dirtly injectors that don't atomize well would cause a mileage drop in the same way that a dirty fuel filter would.


Thoughts / comments ?