Tegger's real-world oil consumption

If someone says that for the last 42000 miles they have added a quart of oil at 1600 miles and when they go to change the oil at 3000 miles the engine is almost down to the add mark on the dipstick, I would then surmise that engine uses a quart of oil about every 1600 miles. When I see this chart I surmise that this guy has managed to create a method that maximizes the uncertainty in what should be a fairly simple thing to keep track of. His engine uses maybe a quart every 1400 miles or maybe it's 1700 miles to a quart. Who knows?. That is pretty sad to go to all that extra trouble and end up knowing less as a result. It's like an old farmer whose method to determine how many cows he has is to count all the legs and tails he sees in a field and divide by 5. You may think that method is absolutely brilliant and I respond that it is utterly ridiculous. The reality for the farmer who has quite a few cows is he never really is sure how many cows are in the pasture.

-jim

I would be suspicious of anyone claiming that they knew what their oil consumption is within 5% over a 42k mile interval. Why? Because driving conditions likely play a role as much as measurement error.

I think folks who are not as careful as Tegger should be prepared for variations in measurement of around 25%. They should not read anything into supposed increases of say 25% until they have monitored oil consumption over a long period.

Sure a study could always be done better. But this one is in fact very well done. Failure to concede this subtracts from a person's credibility, in my eyes.

There is talk on the net of sudden increases in oil consumption, and then consumption reduces by a lot. But such observations may in fact be due to mere measurement error. I think this study helps support this theory.

You and I came to the same conclusion about what the study said, per the quotations in my last post.

This study and subsequent reporting may be careful. Nobody said it wasn't careful. You can take a s*it very carefully, but you still have nothing but turds to show for it.

The variations of 25% are a direct result of the invalid methodology used. If you simply figure out how much the engine used over the entire oil change interval (3000 miles in this case) you will get a reliable result. If you use 1.8 quarts in 3000 miles traveled that is 1670mi/qt. If you use 2 quarts in 3200 miles that would be 1600 mi/qt. Simply accounting for all the oil used and all the miles traveled and dividing will give an accurate result and eliminate the wild variations. Doing it the way shown on the PDF file is lot of extra work that does nothing but make the result less accurate and more uncertain.

It is just plain stupid to do it by measuring small increments on the dipstick over varying mile intervals and then averaging the results of those mi./qt. calculations. To illustrate the stupidity - let's suppose he had read his dipstick at intervals of 200-400 miles and calculated the mi/qt for each of those varying mileage intervals and then averaged those mi/qt calculations. You may call that being more careful, but would it be more accurate? No it would not! It would produce an even less accurate result. I suggest you enroll in some remedial Math education if you can't understand why that methodology is invalid.

That's very funny.

WTF??? You are now claiming that the value of this study is that it is an excellent example of how one can introduce measurement error?? What purpose does that serve?

I echo what Dave Kelsen said.

Hey Jim,

You are dissing a guy who has provided good information for me and others for quite a while here. Lay off of him. He does more work to make this newsgroup valuable than just about anyone else.

Tegger,

Please don't be discouraged by his dissing you. Every newsgoup has those who want to criticize without offering their own work for review.

I appreciate you time and effort here. I know there are many more who do.

Michael

Another person heard from who is stupid and proud of it.

I responded to that because it seemed to indicate he actually was interested in why this methodology might be flawed.

The problem is actually much worse than just "the issue of measurement accuracy when reading off the dipstick". Besides the measurements just being slightly wrong, there are two additional problems that may compound or significantly magnify the initial measurement errors.

A- drive 1000 mi .5 quarts down on the dipstick = 2000 mi/qt B- drive 1500 mi 1.2 quart down on the dipstick = 1250 mi/qt -------------------------------------------------------- Average of A and B = 1625 mi/qt

Now even if we pretend the measurements were dead nuts accurate what we have is 1.7 quarts used in 2500 miles which comes to 1470 miles per quart as the actual oil consumption. The 1625 is just a bogus number.

Calculating how much oil an engine uses isn't rocket science and certainly doesn't require complicated charts and complex calculations. I once owned a car that burned 600 miles per quart. I drove it for 250,000 miles and it used the same amount of oil during the whole time I had the car. I could predict within 20 miles when the dipstick would exactly reach the add line. It went about 800 miles on the first quart, then 600 miles thereafter. This was because when the oil was changed 5 quarts brought it to slightly above the full line. Adding oil when it hit the add mark brought it up to slightly below the full mark. Just like clockwork over and over that pattern repeated. I used a little thicker oil in summer but it didn't make a noticeable difference in the consumption pattern.

jim wrote in news:07idnZY6y6X0_5fRnZ2dnUVZ snipped-for-privacy@bright.net:

That's a good point, and one I hadn't explicitly considered.

I have noticed, though, that if the stick shows, say, 0.8 low on my paper chart and I add 0.8, the level goes back up to its position during the first check for that 1,00 miles. The same thing happens if the level is down 0.4 on the chart. I therefore had no reason to suspect that the level might not decline linearly.

But since it's a possible variable that I need to account for (now being aware of its possible existence), this is something I'm going to have to confirm.

Elle wrote in news:f460cffc-05c7-4f44-832d- snipped-for-privacy@c33g2000yqm.googlegroups.com:

I didn't, unfortunately. See my reply to "jim".

I put the fresh oil into a graduated container and pour it into the engine from that. This way I can make certain I add exactly what the engine used. If my measuring and chart are correct, then a reading of x-milliliters low low ought to be exactly offset by the same amount added back in. And it is, so far as I can see.

jim wrote in news:8bydnYVqdqkwGZfRnZ2dnUVZ snipped-for-privacy@bright.net:

It was left out because I didn't think to put it in.

Part of my reasoning for posting this to Usenet and BITOG was to solicit others' opinions on the test, its methodology, and the report. These I have received (some a bit irascibly), so thanks for that.

I have updated the PDF to account for the issues that have been brought up here and in BITOG.

Elle wrote in news: snipped-for-privacy@r27g2000yqb.googlegroups.com:

I recorded everything I found as precisely as possible. Nothing is "missing" from the dataset.

However, one record was questionable and was therefore not included: my very first properly-recorded check gave me a mileage of 2,200. This created an anomalous spike (that did not occur at any other point in the test). I therefore decided I'd done something wrong during that particular check, and excluded it from the results.

Michael wrote in news:NHPOn.356264$ snipped-for-privacy@en-nntp-01.dc.easynews.com:

Thanks.

But "jim" actually did have a couple of good points buried within his invective. I've adjusted the PDF to address those points. And I'm going to see if I can determine exactly what is the degree of linearity of the dipstick readings.

jim wrote in news:VOednbSH9Z--aJbRnZ2dnUVZ snipped-for-privacy@bright.net:

An initial check is taken before each test sequence.

But you're only using two data points. I suspect that, as the dataset grows ever larger, that the difference between your first method and your second will lessen greatly, and will eventually disappear. That's why sample-size is so critical to any sort of statistics.

This methodology means that there is not a need to do a linearity check, at least for this study.

I quoted what you wrote above at honda-tech.com . The threads are in the tech/misc and Acura Integra sections.

Aside, for those trying to post to the BITOG forum: I applied to join the BITOG forum almost a week ago and still have not been approved.

The point I have been making is your PDF file is itself strong evidence that your measurements and calculations are flawed. I can tell you there is very little doubt that the error exists. I would estimate there is a one in a million chance that the measurements and subsequent calculations shown in the PDF are at all close to correct. I believe the oil is being consumed at a fairly steady rate. The wild fluctuations are nothing but evidence that the measurements and calculations are inaccurate. The fact that the fluctuations are not at all random suggests that the error is due to a measurement bias. The bias makes it appear that typically more oil is being consumed in the first 1000 than in the second 1000. But in reality the general difference that is found the first and second readings is probably due to something other than the actual consumption. Whether you are interested in finding the source of the error is your call. no one else is in a position to investigate. On the other hand you can circle the wagons and try to defend the foolishness. It's your call.

One thing to consider is oil at 2500 miles looks different than fresh oil. And your measurements and calculation entirely depend on how the oil looks on the dipstick, so that is one possible source of a measurement bias.

The suspicion that the error comes from the irregular shape of the pan was based on the assumptions made at that time (which I posted). There could be a small region on the dipstick where there is a non linearity. If your first 1000 mile measurement tends to be below that spot and the second above then you could be getting large difference where in reality both measurements are close to the same.

I will state my current assumptions based on some new info in case you're interested:

Notice that I used 1000 mile intervals but your PDF file makes it clear the 1000 mile intervals is most of the time just a fiction. Also, there should have been 43 dipstick reading in 42000 miles, but only 25 of these were recorded. Did you not like what you saw on the dipstick on those other occasions? n other words. what bias was introduced by selecting only 40% of the possible samples. What about the

-jim

No sorry doesn't at all work that way. You claim to be trying to determine how much oil is being used on average. Your method arriving at that number is grossly unreliable. However your data is too spotty to actually estimate how inaccurate that method is for this data.

The correct method is easy. If you summed how much make up oil you added in total plus how much less than full it was at the time of oil change, you would get a number that represents the total consumption over the entire 42000 miles. You then make the calculation on that total consumption and total miles.

Try your method with the IRS. Tell them you are going to average dollars/day for various periods of income that you selected by some unknown criteria and in which some of them you earned a lot of money per day and some periods not so much per day and that you will use that average of those 17 dollars/day figures as an accurate measure of your annual income. The idea is so absurd that you probably will even get an IRS agent to laugh. You don't have a statistical problem to solve - you have an accounting problem and you are applying absurd accounting practices.

-jim

Measuring engine oil consumption in this manner is difficult. While you definitely can measure the volume of fluid in the crankcase, that way not mean much unless you also know the make-up of the fluid in the crankcase.

All piston engines consume some oil. They must. A thin layer of oil remains on the cylinder wall on every piston stroke. Some of this is burned during the combsution process, some vaporizes and leaves with the exhaust, etc. Even in a very well sealed engine, more oil leaves via the valve stems, the crank seals, through the PCV system, etc. More of the original oil leaves as light components of the oil boil away.

Balancing the oil that is consumed is comtamination added to the crankcase via blow-by past the rings, through the PCV system, air exchanged with the atmosphere, and in some cases minor leakage from the cooling system. This contamination takes several forms - water, unburned hydrocarbons, soot, dust particles, etc. Some of this contamination ends up in the oil filter, but much of it remains suspended in the "oil" in the crankcase. Some of these contaminates have realtively low boiling points and can be driven off if you get the "oil" hot enough. Others don't boil off. The fact that the oil changes color is evidence that some contamination remians.

When you first change your oil, you have a crankcase full of pure oil. A thousand miles later, the crankcase is now filled with a mixture of oil, water, soot, unburned hydrocarbons, dust and ? If you check the oil level immeadiately after a long drive which has rasied the temperature of the oil above the boiling point of water, you can assume that most of the water, and much of the unburned hydrocarbons have been vaporized and are not immeadiately contributing to the oil level in the crankcase. But soot, some unburned hydrocarbons, and dust are still present. It seems reasonable to assume these are of only minor significance, and that the decrease in oil level really does represent oil consumed. However, if you check the oil after several days of short trips in cooler weather, it is likely there is still significant volumes of water and unburned hydrocarbons in the crankcase contributing to the oil level. I am not sure how you can manage the measuement process to consistently guarantee that you have driven off the water and unburned hydrocarbons to the same extent?

You might think this is a trivial problem. I don't. On my farm we had a very old gas tractor. The combustion rings did not seal well and the tractor suffered from massive blow-by. We typically just used the tractor to pull a trailer, or move a scoop load of grain for the cows (it had a front end loader). Much of the time the oil level in the crankcase stayed level or actually increased. Occasionally we would work the tractor harder (for instance using the loader to move a pile of dirt). In these cases, the oil level would drop dramitically. I never bothered to have the oil analysed but I suspect that after just a few months of light usage it included a very high percentages of water and unburned hydrocarbons.

Another factor would be when and under what conditions you add make-up oil. The rate of oil consumption should increase as the composition of the oil in the crankcase changes. I believe that the "fresh" oil probably has a realtively low given burn off rate. As the composition of the oil changes, this burn off rate changes as well. What is left of the original oil is likely more viscous and likely burns off at a lower rate. However, the contamination that has accumulated in the oil probably has a much higher burn off rate. So for a engine with weak / worn rings and seals, the percentage of "original" oil steadily decreases but at decreasing rate. I am guessing that you might see a realtively quick decrease initially, followed by a long slow increase in usage that levels out, asusming you add fresh oil regularly to compensate for usage. At some point you are replacing contaminated original oil with fresh oil in such a way that you reach a sort of steady state composition (and burn off rate) for the oil.

Ed

Elle wrote in news: snipped-for-privacy@u26g2000yqu.googlegroups.com:

Try applying again.

jim wrote in news:vLCdnVF7yL_FjJDRnZ2dnUVZ snipped-for-privacy@bright.net:

You make some fair points.

I respond by posting my raw data. They are in an Excel file, here:

The first two columns give the actual miles driven during each test, and the actual observed amount of oil used during that test. (It must be noted that, in order to maintain consistency with the "reported" mileage, the oil amounts are slightly adjusted. i.e.: 0.622 is actually 0.6; 0.597 is also actually 0.6.)

I know you chose extremes in your example in order to make a point, but now, having /real/ numbers to work with, you do the arithmetic, and tell me what you get.

Remember that doing it my way, I get 1663 mi/qt. Doing it your way gets...what number?