M42 Fuel Economy

[Bekverdy30]

Hello everyone, I just had a quick question on fuel economy in our cars. My car is a stock 1991 318i Convertible and I recently noticed how bad my fuel economy is. My car has about 160k miles on it, no oils/ vacuum leaks, and it gets about 18-20 mpg. I pump gas about once a week when I have about 3.5-4 gallons in the tank. And in that one week interval, I drive about 180-200 miles before I have to fill up. Meaning I am getting around 18-20 mpg when I have read on other forum posts about people getting 27-30 mpg with cars that have ALOT more mileage than my car. Also, I pump 91 octane Shell gas, which is not cheap around where I live (Los Angeles), and I do not drive my car hard since I am sitting in traffic most of the time. But even when sitting in traffic, I do not think that should kill my mpg that much. My questions are, what could possible issues be and what do you guys normally get?

Things I have changed that could come into play:
-Fuel pump
-Fuel filter
-Spark Plugs
-Air Filter
-Upper and Lower intake manifold gaskets
-"Mess under the intake" delete

[DesktopDave]

That honestly does seem a little low. Granted, the cabrio is a lot heavier, idling in traffic is expensive for fuel, and A/C is a gas hog. Have you checked your coolant & intake temp sensors? How old is the O2 sensor?

Overall my averages are 25.1MPG city (suburbs) and 35.1MPG on the highway. I'm down a little since the car usually gets cheap gas and I went back to 10w40 oil.

TL;DR: Basically the only thing that really worked to improve fuel mileage was "Top Tier" 93 octane fuel, the hose delete mod, 4-hole fuel injectors and really thin 5w20 motor oil.

The COP kit, P/S delete, new spark plugs and diluted coolant (66% water) also helped a little. I also dropped the diff ratio from 4.1 down to 3.73 but I have no idea if that helped since it was the first thing I did when I got the car. I also tested a larger 17" wheel & tire combo, (255/45 R17?) that significantly cut both fuel mileage and ride quality. It handled beautifully though.

[Bekverdy30]

I have changed both coolant temp sensors (the one underneath the intake manifold and the one on the radiator). I have a brand new rebuild AFM that has less than 1000 miles on it. The O2 sensors have also been changed around 18k miles ago before I bought the car. I have the 4.27 open diff ,which sucks on the highway, and I run 10W-40 Castrol GTX motor oil. I'll check the gap on my plugs as well to make sure that might not be the case.

[DesktopDave]

Forgot to add in my previous post...I wouldn't suggest running thinner oil. I loved the higher mileage but my used oil reports had very high iron & chrome readings indicating insufficient piston lubrication. I'm keeping it at BMW's recommended weights depending on temperature for now. I am trying out some of the high-mileage formulations, I'll report on that once I get some more miles on the car.

GTX is a great choice, I used that for years. You might pick up a little with synthetics in the engine, transmission & diff but it's unlikely that you'll gain enough to pay for the higher cost unless you run really high intervals like I do. I tend to go about a year between filters & top-ups...10-15K miles between oil changes are OK based on my UOAs.

 I've heard great things about RedLine. I also really like the MolyLube LiquiMoly products, although IMHO they're too pricey for our humble little Bimmers. On a (possibly) related note, I can highly recommend Ford Motorsport XT-M5-QS gear lube. It's a dedicated MTF that massively helped with shifting notchiness in my brother's NC Miata 6-speed. I'm unsure if it's a bit thin for our old Getrags or has other compatibility issues with brass synchros.

[monty23psk]

From what I have seen over the 15 years plus on the internet is that most of the numbers are calculated incorrectly or they are fluff. Dave's numbers seem to be accurate but like everything, variables of city versus highway and they type of driving you do will affect the numbers. The stock numbers from BMW for the 91 m42 from EPA is 18 in the city and 25 in the highway. You can check the EPA website. Also, what gas tank you have also depend as most of them are the 13.1 gallon tank but I have heard some might have the larger 16 gallon tank or smaller 11 gallon one. Also, if you have the 13 gallon tank, you MPG would be lower.

Best thing you can do is get an app that measure how many gallons you add and the mileage and after 3 months, you have a realistic baseline. This would be city or combined. Then after, you can do an only highway trip to see what you get. AC on these cars, because of drag, really kill the MPG.

[DesktopDave]

+1...most numbers are calculated with fill-ups and trip odometers. There's a lot of room for error. Although I was fairly diligent, there's an odometer inaccuracy in my data due to the taller diff and tires.

I wish I had a 13 gallon fuel tank...the PO bounced it off a few things. Even filled to the very top of the fuel neck I've never gotten more than 12.7 gallons into it. One of these years I'll re-inflate it, I guess.

I also think it's easy to forget the occasional top-up, which would significantly increase observed fuel mileage. Although I use the trip odo as well,
my spreadsheet doesn't have that problem...I use the main odometer as an error-check, so it doesn't matter if the trip odo was reset accidentally, someone else put fuel in the car, etc..

[timothymcn]

Idiot here. Why would the 13 gallon tank get worse mileage? Wouldn't the smaller tank get better mileage, just because of weight?

There are a number of oil threads to look through, but the consensus is Mobil 1 15w50. Best zddp levels for our particular engine. Walmart also has it at a good price. I really don't think using the wrong oil is worth saving a few bucks on gas.

I see a lot of very high mpg numbers on the forums, but I personally get a very consistent 22-23 mpg mixed use (75% City) driving with a heavy foot.

Also - I would guess the drag isn't actually as bad as you would think from the boxy shape. Drag coefficient is only a part of the equation, and if you can trust old  magazine numbers, .33 is the same as many modern cars, combine that with a smaller frontal area compared to many modern vehicles, and I would guess drag isn't a big factor.

[DesktopDave]

Idiot here. Why would the 13 gallon tank get worse mileage? Wouldn't the smaller tank get better mileage, just because of weight?

I think he's noting that guys with the larger fuel tank would be posting a lot more miles per fill-up, not saying a larger tank would get better mileage. 

There are a number of oil threads to look through, but the consensus is Mobil 1 15w50. Best zddp levels for our particular engine. Walmart also has it at a good price. I really don't think using the wrong oil is worth saving a few bucks on gas.

+1 Hard to beat WallyWorld's price on oil, although I tend to have trouble finding thicker oils at my local store. Their SuperTech synthetic is pretty good too - it's even cheaper - and gets pretty good reviews over at BITOG. Mobil1 has been excellent for me even up to 12K oil change intervals. I use it in all my vehicles unless I've done recent engine work. For breaking in an engine I'll use whatever brand-name cheap/thick 15w50 conventional oil is on sale.

My VFR750 is a special case...motorcycle synthetic oils tend to be expensive and they require pretty frequent changes. Since my wife's Kia is still under warranty the oil has to be changed every 7.5K miles. This might sound a little crazy, but I'll put one quart of new 10w40 synthetic, then add three quarts of used oil 5w20 from the Kia. The thinner oil helps out with cooling (VFRs run notoriously hot) and the 'broken-in' used oil gives a lot better clutch lever 'bite'. The VFR's wet clutch tends to slip with high-efficiency synthetics and the bike tends to contaminate any motor oil pretty quickly...so I wouldn't push any oil over 2-3K miles. Motorcycles in general are awful for oil. Condensed water gets into both oil & fuel, then add plenty of unburned fuel from the quartet of oversized carbs, then add plenty of debris from the spur-cut dog gear transmission and organic clutch pack, top it off with my tendency to make short trips. Nasty little environment for a lube!

I see a lot of very high mpg numbers on the forums, but I personally get a very consistent 22-23 mpg mixed use (75% City) driving with a heavy foot.

Also - I would guess the drag isn't actually as bad as you would think from the boxy shape. Drag coefficient is only a part of the equation, and if you can trust old  magazine numbers, .33 is the same as many modern cars, combine that with a smaller frontal area compared to many modern vehicles, and I would guess drag isn't a big factor.

The E30 does have a very small frontal area, so the chassis is pretty efficient. The early E30s were all about fuel mileage. The EFI M10B18 was spectacular on fuel, the M20B27 Eta was excellent. Europeans even got the M21 diesel. The M42 wasn't really developed for efficiency IMHO, seems to me it was built for durability and high-RPM autobahn driving. The M44 was a slightly more efficient design, albeit a little less fun.

To make things worse, our cars are frankly under-powered for US roads. To 'fix' that problem, BMW fitted very short diffs to US cars. Both high-RPM highway use and stop-and-go commuting are very hard on fuel mileage. Speaking of that, my biggest overall difference in fuel mileage is my driving style. I tend to run my car very easily; lots of highway miles, most starts in 2nd, coasting up to stops, etc.. I don't get nearly that kind of mileage in my two-ton 635CSi. I have a tendency to flog that car more than I should. The high-compression M30B34 is a hoot to drive but it's hog on fuel!

[bmwman91]

(all of my fuel economy mentioned below is based on actual trip odometer miles divided by pumped gallons for a full tank)

I have been noticing a consistent decrease in fuel economy in my car over the last decade or so. This was both with the original internally stock M42, and my 2.1L built M42. It has perplexed me to no end, and I have replaced basically everything that I can think of which might be causing this: everything in the fuel delivery system, O2 sensors, ECU tunes, wheel bearings, rear diff, driveshaft & consumables, etc. None of this has helped.

Back in 2008, I was getting 29-33MPG commuting in mixed city/highway settings. Granted, because of the bad traffic, I turned it into a game of fuel economy optimization, with a lot of coasting and such. Still, 29MPG was basically the default when I was not really trying. Fast forward to ~2011. The same engine was now getting more like 23-25MPG. Major changes included me moving MUCH closer to my office, so the car was largely used for short trips in town, bu even still, my highway mileage took a dive too. Then that engine died, and I had the 2.1L built. Ever since then, I consistently get 18-21MPG with mostly city driving. Some of that is because this engine is a ton of fun, and I redline it every chance I get. I did do a 250 mile drive to visit a friend a couple of years ago, and I decided to try to get the best possible mileage. No flooring the gas, shifting at 4000RPM or less, and going exactly 65MPH. That netted me 26.5MPG, and that is the best I have ever managed with this engine.

As far as I can tell, one big change here in CA around 2010 was a bunch of new air quality regulations that mainly resulted in adding lots of ethanol and other oxygenators to gasoline. I think that most CA gasoline has ~10% ethanol by volume. That does not account for the ~15% decrease in fuel economy (for "behaving myself on the highway" cases), but it explains a lot of it. Other states have probably had some number of changes in the last decade, too. I need to see if some other county has different requirements and if I can try filling up with gasoline with no ethanol content. That is the only thing I have not yet tried in order to verify the cause of my lousy fuel economy.

[DesktopDave]

We've had some similar changes up here. The ~15% ethanol blends are great for raising octane...but awful for fuel mileage. Alcohol is only worth what, like 85% of gasoline's BTUs?

To make things worse, up here in the (formerly) great white north, they'll also change the fuel blends when the weather cools off. Butane is added to lower the flash point so cold starts are easier. In an absolute sense, based on my personal data, winter blend fuels cut mileage ~5%. Another way to think about it...if we'd prohibit carbureted vehicle use during the cold months the whole state could use 5% less fuel.

[monty23psk]

I did forget to add the gas being used has changed due to ethanol and how it burns. Also, in between brands makes a difference. Our cars were engineered for a different gas, so the spark/burn has different characteristics if something changed.

in the end, if wallet is not an issue, is more about SPGs than MPGs when you drive. SPG = Smiles per gallon

[bmwman91]

Let's see...15% by volume and ethanol has around 2/3 of the energy density. So there is a 5% hit right there, versus straight gasoline.

I also wonder how it affects the O2 sensor feedback system. Ethanol will cause a leaner mixture than gasoline for a given base fuel injection value. The O2 sensor will tell the ECU to compensate for the leaner readings and then increase the injection times until it sees a 14.7 ratio. Stoichiometric for ethanol is 9:1. It is injected at a 14.7:1 ratio, which will mean a 5% increase in fuel injection mass (15% ethanol content x 33% extra oxygen from ethanol's lower AFR). Densities of gaoline and ethanol are a bit different, but a 5% increase in injected fuel mass means a little more than 5% more volume injected.

So those two things there mean a 10% hit to fuel economy. God only knows what the effects of the winter blend stuff are. I assume that other factors are also at play here, which don't help our situation.

[monty23psk]

Not sure how much of this is accurate:  https://www.pure-gas.org/

[bmwman91]

Interesting. Looks like I may have one option in my area. Assuming I can find the time, I will see about running a few tanks of straight gasoline to see how it changes my mileage, and if the car "feels" better on it.

[monty23psk]

I will do the same when I have my care back on the road. Gold mine if we find some and other reports online, including this site, explain, as you stated, the ethanol is not meant for older cars, and degrades the performance. I wonder what year is the cutoff, when ethanol was introduced? I haven't had time to investigate.