By on June 18, 2013

TTAC commentator wannabewannabe writes:

Hey Sajeev,

I wrote you a while back about my 1990 Chevy pickup, but I’ve since moved on to more interesting cars.

I have a 1994 Cadillac Fleetwood Brougham (B/D-Body with an LT1 FTW!) that looks literally identical to this one, same color inside and out. It currently has just under 184k miles on it, is my daily driver, and is in pretty good shape. Very reliable car!

My question concerns fuel economy, and it’s stumped me. I commute between 75-80 miles round trip every day, which makes my weekly mileage about 400 miles, so I fill up once a week. I always fill up at the same gas station, I drive the same routes on the same days, no one else drives my car, and because my commute is mostly freeway, I’m a copious user of cruise control. Also, I should mention that because I live in the Bay Area in California, our weather doesn’t vary wildly; it’s usually relatively mild. Yet given all this, my gas mileage routine varies up to 15%. Some weeks I average about 21mpg, while other weeks I average about 24mpg. To be fair, the average is in the 22.5mpg range, and you’ll see a good number of data points there too.

I would expect some variation based on the automatic shut off on the pumps and to a certain degree, variances in temperature, particularly when filling up, but I’m surprised that my fuel economy swings as much as it does. I’ve included a rudimentary chart to show you the swings, though it doesn’t have the last month of data included. Thoughts?

Sajeev answers:

Nice rig!  You’ve certainly done a great job explaining your situation..and is it ever a doosie!

The most logical choice, given your input, is the gas pumps do shut off at different “full” levels. Perhaps there’s something wrong with your Caddy’s vapor recovery system that’s making it happen.  Usually that throws a trouble code and a light on the dash, so probably not applicable here.

Perhaps the weather may feel about the same, but the problem is exacerbated by a change in barometric pressure? Is that even possible?

If, by any chance, you haven’t changed your Oxygen (o2) Sensors, do it!  The engine computer is only as efficient as the information it receives.  And flaky sensors could cause the variance we see here, especially if they are original at this mileage. Worn out, but not enough to throw a trouble code. It’s possible!

The O2 sensors sound more likely, the more I ponder this. Off to you, Best and Brightest!

 

Send your queries to sajeev@thetruthaboutcars.com. Spare no details and ask for a speedy resolution if you’re in a hurry…but be realistic, and use your make/model specific forums instead of TTAC for more timely advice.

Get the latest TTAC e-Newsletter!

35 Comments on “Piston Slap: Charting the Changes, To What End?...”


  • avatar
    danio3834

    Not bad mileage out of that barge, I’d say you don’t have much to complain about. I certainly wouldn’t pound O2 sensors into it based on this info. I would chalk it up to the fact that no two trips are the same, even on the same route. Sometimes you’ll idle a little more on a given cycle, maybe accelerate a little harder, stop a little shorter. Many variables. This isn’t something I’d sweat.

    For a time I drove a baby LT1 (4.3L V8) Caprice daily and would typically average 23 mpg. Sometimes a little more, sometimes a little less. You’re in the ball park for sure.

    • 0 avatar
      wannabewannabe

      I agree. The overall mileage isn’t bad, and probably pretty close to what you would expect from the car, based on my mostly urban freeway driving (the best I’ve ever seen out of it on a roadtrip was 26mpg). I’m just mostly curious about the variation.

  • avatar
    TR4

    I wouldn’t bother comparing one tankful’s mileage with the previous or next one since there is too much variation. Depending on the design of the tank and filler pipe, it is possible to get air trapped sometimes and not be able to fill the tank up completely. Also, due to the flat shape of many tanks a slight difference in the levelness of the vehicle (weight in trunk, slope of the ground at the pump etc.)when filling can make a difference. You need to average the mileage over perhaps three to ten tankfuls to get meaningful data.

  • avatar
    greaseyknight

    I’m not very familiar with that GM system, but its an OBDI not OBDII which has much more advanced controls on evap and emissions and such. So something could be wrong with the evap system and the computer does not know about it. Your fuel economy could also be tied into the amount of Ethanol the gas has on that particular week.

    • 0 avatar

      Good points.

    • 0 avatar
      wannabewannabe

      I thought briefly about the ethanol content, and besides it being E10 basically everywhere in California, is there anyway to really know what the actual ethanol content is and whether it changes to any significant degree from week to week?

      As a side note, I’ve noticed that with all the cars I’ve owned in California, I always make worse fuel economy on California gas than I do on gas from other states, usually about 10% worse.

      • 0 avatar
        Scoutdude

        There will be no significant difference in the amount of ethanol no matter where you purchase gas at in CA. On the one end you have CARB mandating a minimum of 10% ethanol and at the other the EPA mandating no more than 10% so the range would vary from say 9.95 to 10.05 at the most lest the producer face fines when those agencies pull a random sample. Not sure how often the EPA does it but CARB is pretty strict about regular sampling. And yes for many cars E10 does cause a loss in MPG of around 10%.

  • avatar
    Felix Hoenikker

    Wannabee^2

    Since you have the data, you could do a control chart to establish the upper and lower limits of the “process” which in this case is fuel consumption. I did this for a similar situation to yours except that my route was mostly local driving with no highway speeds. Look up control charts on google and you can see how this is done.
    I can share my results with you so you might not feel the need to go that far. I found a similar variation to yours (+/- 1.5 mpg) with a mean about 23 mpg over about three years of driving. Since I live in the NE there as a seasonal variation in average mpg too, but the variation in mpg was a constant +/- 1.5 mpg. The only cause that I could find for this variation was the variability in the shut off in the gas pumps.

  • avatar
    optixtruf

    You need to establish a “control” gas pump to eliminate the gas pump shut off variable-or at least, pay attention to put exactly the same amount in, every time (as, legally, all gas pumps within a state have to be calibrated the same). You’ll see a seasonal change for sure, on certain vehicles I’ve seen it up to 3 mpg difference between winter vs. summer blends. Your o2 sensors are a good start, another thing to consider may be making sure the MAP sensor (Fairly sure that engine was TBI, no maf?) is functioning correctly to respond to the baro change. Also, as hopefully this is obvious, monitor the ratio of city/hwy. Last but not least, You may want to check the operation of the egr system, as over time these may clog (especially in the transfer tube and intake connection) with carbon and affect the running condition; these early onboard diagnostic systems likely had high thresholds for setting fault codes.

    • 0 avatar

      Good info. It’s port EFI…and I am 99% sure LT-1s had a MAF.

    • 0 avatar
      wannabewannabe

      Wouldn’t a malfunctioning MAF or clogged EGR throw up the check engine light?

      As for gas pump control, the gas station is popular enough that it’s almost impossible to secure the same gas pump every time, particularly when I’m in a rush trying to get to work, which is pretty much always. To try to counteract some of the variance in the automatic shutoff, I usually pump one click past the first automatic shutoff.

      • 0 avatar
        danio3834

        “Wouldn’t a malfunctioning MAF or clogged EGR throw up the check engine light?”

        If they’re outside their fault tolerance range, yes. A dirty MAF could affect performance to a point but not past a fault tolerance range. The EGR, if carboned up and stuck could affect performance but will typically set a light if the problem is with the valve itself. Yes the tubes can carbon up too.

        But those issues are mostly irrelevant to your question, as both issues would represent a consistent degradation in performance, where you find a variance up and down. This tells me your deviation is due to normal variance factors with your driving or tracking methodology.

        Honestly, as I’ve told people with similar questions regarding such minutia when diagnosing their fuel economy issues issues after everything checks out, your mileage may vary.

        • 0 avatar
          optixtruf

          Agreed. I track most long term trends going on with my jeep, including fuel econ (I use the same pump every time, and wait to use the same pump), OBD2 readings (mostly fuel trims, throttle positions, spark advance, load, baro, and various other sensors to determine either a degredation in vehicle or sensor performance), and oil analysis records. these work well to see how my Cherokee is aging and what components will soon need work (I believe in proactive, not reactive), but do diddly squat for different driving habits.

  • avatar
    morbo

    You only have 17 data points in your graph. While good for gross level trend analysis, i would wait until you have 30+ data points before drawing any conclusions. Even then, the assumption of the data points being under identical conditions (temp/fill level/acc-dec/tire pressure/etc.) isn’t truly valid. 17 Data points is insufficient for statistical significance.

    • 0 avatar
      wannabewannabe

      I have more data points, but I moved in September of last year. Thus, my commute changed. It basically doubled, but I was able to change my work schedule slightly so I avoid traffic better. Before September, I was driving about 35 miles round trip each day and averaging just under 20 mpg per tank. With the longer commute, I’m at 76 miles round trip and averaging just over 22 mpg. I didn’t include the earlier data points because you would have seen the jump up when I moved, which would make the chart appear less relevant. Different commute aside, though, I did see about the same variation before moving as I do now. Then I would see 18-21 mpg and now it’s 21-24.

      • 0 avatar
        CoreyDL

        Your commute is nuts! 76 miles a day?! I do 6 miles a day RT and I feel like that’s plenty, ha.

        Be happy you get this sort of mileage – you’re besting my GS430. Best I’ve ever seen is 23 out of a tank, when I spent 90% of it on the highway. I get 18-19 in town usually, with no AC. Dips to 17 in the winter.

  • avatar
    dastanley

    I remember as a kid in the 70s, my grandfather would keep a notebook in his glovebox with gallons per fillup and miles traveled. He was meticulous about calculating mileage for his ’73 Pontiac LeMans. Since he only averaged in the lower teens for mileage with mostly highway driving (350/350 4 bbl), I always wondered what his point was in going through all that – his car just used alot of dang fuel. But to each his own.

    For a 19 year old Caddy to be getting in the low 20s combined city/hwy, I’d say you’re doing well and shouldn’t worry about it. Since I’m assuming the car is long since paid off, I’d just enjoy the “free” transportation for as long as you can and realize that chasing this “problem” will probably only frustrate you and yield diminishing returns.

    • 0 avatar
      wannabewannabe

      I’m not too worried about the fuel economy overall because the long term average remains pretty constant. Like your grandfather, I keep a logbook in the car and note all my fill ups: date, location, total miles on the odometer, price per gallon, gallons, and mpg. Because I keep the logbook, I notice the variation more acutely than if I just filled up and drove. I try to control for the variables as much as I can in analyzing the numbers (though I’m definitely no statistician), and I was curious what the B&B would make of the variation.

      • 0 avatar
        dastanley

        As an air ambulance pilot, I get paperworked to death in every aspect of the job. So as far as my cars go, I just gas and go out of sheer laziness and data fatigue. But I can certainly understand you keeping meticulous records and the value you assign to that data if it helps you spot trends, etc. I know my ’06 Corolla (I know, I know…) gets somewhere in the 30s and my wife’s ’08 Tucson gets 20 mpg or so (V-6 AWD). That’s enough for me to know.

        • 0 avatar
          wannabewannabe

          Sounds like the Tucson makes roughly the same gas mileage as my Caddy, which seems kind of ridiculous for 14 years of progress… How much does the Tucson weigh? The Caddy’s vital stats: 5.7 liter V8 and 4400 lbs. curb weight.

          • 0 avatar
            galanwilliams

            I have been using an app to track my mileage / fuel cost — provides a lot of interesting statistics — but in the 14 months that I’ve been recording the mileage of my 2005 Jeep Liberty 3.7L 6-cyl, 4-speed auto, which weighs 3900 curb, my average economy is 18.6 mpg, with a best and worst of 24 and 16 — so clearly the tank of 24 mpg must have been mostly downhill. :) I too would think that it should be better, especially as I try to drive judiciously – and this is at a cruise speed of 65 – 70… anything above 70 cuts the MPG by another 2 or 3!

  • avatar

    It’s quite possible that your variation is simply within the error margin of your methodology. As mentioned, you might not be accurately measuring how much fuel you’ve used (you only really know what you’ve put in there). Further, in warmer weather you’ll get more evaporation during filling (and fuel is denser in cooler weather) so filling up in the night or early morning might result in more fuel – and thus apparently greater fuel usage – from your prior tankful, than if you filled up in the heat of the day in, say, the late afternoon.

    Also, your driving is never going to be identical. Air conditioning usage will cost you fuel. Driving with open windows at speed might reduce fuel economy. Stop and go traffic will reduce your economy, as will red lights. The more you brake, the worse your economy, too, since momentum built up using fuel is being turned into heat instead of being used to move the car forward (that’s what braking does).

    I’d be more concerned about your long-term patterns. Figure out your mean and standard deviation, and see if your overall pattern remains the same even if the variability remains. I expect you’ll find that your overall fuel economy remains rather stable over time, within the deviation you’re seeing. If it starts to drop in the long run, that’s evidence of a problem (but don’t forget to consider any seasonal variation you might endure – might not be a big problem there, but where I live, we get our best economy spring and fall and our worst summer and especially winter).

    • 0 avatar

      I’ve found on my last 4,000 mile road trip with my ’77 Chevelle that it backed up my thinking that the big A/C compressor used 3-4mpg at any given speed, and that its sweet spot for fuel economy was 60mph. At a steady 60 with no A/C and the windows up, the Eisenhower era engine and fuel controls got 23mpg. At 75 it got 18mpg.

      Turn the A/C on and it drops to 20mpg, and then to 15, respectively, and in town it gets a pathetic 12mpg with it on.

      • 0 avatar

        Speed – that’s a good point.

        I recently got the best fuel economy ever in my ’07 Accord with 2.4L I4 and 5MT – 6.0 L/100 km – driving 260 km to Saskatoon during a raging blizzard. I rarely got above 90 km/h (speed limit is 110) and was driving more like 60-80 for long stretches, and I had a huge tail wind behind me. That works out to 39 mpUSg. My previous best tankful had been 6.9 L/100 km.

        • 0 avatar
          wannabewannabe

          Speed definitely makes a difference. Though I haven’t experimented with the Caddy much, I have with earlier cars. When I had my 90 Corvette (base coupe with an auto), the absolute best gas mileage I got from a single tank of gas was 30.4 mpg, which was obtained over 400 miles of going about 60 mph from Santa Fe, NM to Steamboat Springs, CO.

    • 0 avatar
      wannabewannabe

      As it’s been a bit warmer lately, I have been using the A/C more, but at least in this car, the A/C basically has no effect on the fuel economy. Benefit of having a big, lazy V8 with extra power, I suppose. In any event, as you can see by the graph, I wrote this question before summer arrived, so A/C wasn’t really an issue through the fall and winter. And just so everyone knows, winter temps here range from about 40-60 and summer is 55-80, so it’s pretty mild most of the year.

      • 0 avatar

        You should always see some effect, unless your fuel economy is so poor that the additional usage of the air conditioner is trivial by comparison. Your economy isn’t that bad.

        This, to me, is evidence that there’s a bit of slop in the measuring (no insult to you; it’s hard to do this accurately). I’d expect in the long run you should notice some difference as you use air conditioning more – the mean should drop even if it’s hard to see any obvious effect on any given tank of gas.

  • avatar
    jhefner

    I drive a similar commute as you in a ’95 Taurus; and after watching it for a year, believe that temperature causes the greatest variation in mileage — 24 mpg when it is in 50s, up to 27 mpg when temperatures are in the 90s. I think it has to do with how long it is running with the choke on (or what works as the choke on a F.I. car; the colder it is, the longer it runs rich.

    To try out this theory, try logging daily high temps as well, and plotting together. Otherwise, as others said, I wouldn’t sweat it; that’s good for a car your age.

  • avatar
    bumpy ii

    Easy way around the pump shut-off issue is to track it by gallons per fill, and mileage between fills. A “fill” is however much you put in; doesn’t have to be a top-off.

    • 0 avatar
      scrappy17

      Can you please elaborate on that? Preferably with an example.

      I am a little confused.

    • 0 avatar
      optixtruf

      This methodology would work if you ran every tank to bone-dry. Unless you want to measure the existing fuel in your tank pre-fill up (by weight is most accurate), this method would be be pretty inaccurate.

    • 0 avatar
      galanwilliams

      Agreed – this is the reason that averaging is essential — one of the biggest variances comes from not being able to refill the tank to exactly the same level each time. One of the joys of my Liberty is that it doesn’t shut off fast enough, so I always loose some fuel down the side of the vehicle and in the parking lot. So as it gets close, I take the nozzle off auto and fill as slowly as I can until it clicks off – for my vehicle, this is really topped off — I can’t fit any more in!

      Just heard that Jeep has agreed to recall my vehicle for the gas tank being behind the axle issue (at the arm twisting of NHTSA) — but I’m sure that will not involve much more than bolting some sort of steel beam behind the tank, so no help for my overfilling issue.

    • 0 avatar
      bumpy ii

      Put 6.7 gallons in, drive for 203 miles, put 9.2 gallons in, drive for 312 miles, put 8.7 gallons in, drive for 120 miles, put 3.3 gallons in, drive for 290 miles, etc. 10 or 15 data points will be enough to filter out whatever extra gas you might have started with.

  • avatar
    SunnyvaleCA

    I also live in the Bay Area and I also keep a log book of mileage (and other car-related odds-and-ends). I think your problem is the automatic shutoff system of gas pumps. I’ve noticed that if I lock the gas pump nozzle to maximum fill speed it usually stops short. Usually it’s between 0.5 and 2 gallons short but sometimes it’ll just stop anywhere. That could entirely explain your inconsistent results.

    While fueling I try to estimate how many gallons I’ll need based on miles traveled, expected fuel efficiency, gas gauge reading, and the car’s own estimation of miles remaining until empty. Then, when I’m 2 gallons shy of full I reduce the nozzle fill speed to the slowest speed and also hold the nozzle more upright. That seems to result in consistent fills.


Back to TopLeave a Reply

You must be logged in to post a comment.

Subscribe without commenting

Recent Comments

New Car Research

Get a Free Dealer Quote

Staff

  • Authors

  • Brendan McAleer, Canada
  • Marcelo De Vasconcellos, Brazil
  • Matthias Gasnier, Australia
  • J & J Sutherland, Canada
  • Tycho de Feyter, China
  • W. Christian 'Mental' Ward, Abu Dhabi
  • Mark Stevenson, Canada
  • Faisal Ali Khan, India