Oil burning test

[natemoen]

So I had some used oil and figured I would see how much gas it took before I could light it on fire.

I was using 5w-30 oil from my wifes car. I started with 1 quart (32 ounces) of oil and started adding 1 ounce of clean new gas at a time to see when I would get it to burn.

As I started adding the gas I would give it a goo mix and then try and light it. As I hit about 4 ounces of gas in the oil I noticed a little difference in how the fire hit the stick but I wouldn't say it was something that would be noticeable if you werent doing this step by step test and seeing how it changed. It wasnt until I hit 8 ounces of gas that I would say that I got what I would call a burn and even at that it was a pretty quick flash as it burnt off the gas quick. When I hit 10 ounces of gas in the oil then I started getting a longer more consitent burn.

So, I would not suggest using this test a a difinitive change or no change for the oil. I would say if it smells like there may be gas in there change it, 8 ounces per quart is way to much gas to have in your oil

[scoot]

Really good to know!

Well done there, Nate! So it's time to retire that old benchmark burn test, and as you say, if it smells like gas, it's got to go. Nice work.

scoot

[cywelchjr]

Yeah, that sure puts that one to rest. If it takes almost 33% gas in the oil to get any significant burn out of it, that's not a reliable test and could give someone a false sense of security thinking bad oil is ok.

[b.walker5]

I might have missed the point of your experiment here Nate, but any gas in the oil is a bad thing. There should be no acceptable threshold, as it's not the oil flashing (burning) that is the problem, it's the oil thinning out and destroying the bearings, because it's not a good enough lubricant under heat and pressure any more.

Don't forget that when inside a moving engine the oil and fuel will be in a super heated blender and mixed together very quickly, and it will take way less than your 8oz to a quart to start damaging bearing surfaces.

[CatatonicBug]

I might have missed the point of your experiment here Nate, but any gas in the oil is a bad thing. There should be no acceptable threshold, as it's not the oil flashing (burning) that is the problem, it's the oil thinning out and destroying the bearings, because it's not a good enough lubricant under heat and pressure any more.

Don't forget that when inside a moving engine the oil and fuel will be in a super heated blender and mixed together very quickly, and it will take way less than your 8oz to a quart to start damaging bearing surfaces.

That is exactly the point of Nate's experiment. Lots of folks suggest dipping a stick in the oil and checking to see if the oil that comes out will light. Thus, determining if there is fuel in the oil. Unfortunately, thanks to Nate's experiment, we now know that this dip-stick approach is not a sufficient test to determine the presence of fuel in the oil.

[SFerinTEXAS]

I believe the moral to this story is........ What CatB said.

[cywelchjr]

I might have missed the point of your experiment here Nate, but any gas in the oil is a bad thing. There should be no acceptable threshold, as it's not the oil flashing (burning) that is the problem, it's the oil thinning out and destroying the bearings, because it's not a good enough lubricant under heat and pressure any more.

Don't forget that when inside a moving engine the oil and fuel will be in a super heated blender and mixed together very quickly, and it will take way less than your 8oz to a quart to start damaging bearing surfaces.

I think that a teaspoon of gas or so in 3 liters of oil (plus the filter) would not be enough to cause damage before it would flash out of the oil because of it's far lower flash point, but you're correct that far far less than 33% gas to oil ratio would be fatal under load (at idle not so much, it would clean out a lot of gunk but if not put under pressure and not run too long would likely not cause damage). That was the point, was to find out how much gas in the oil would cause it to burn on a stick. It turns out that it's far far past the point that would be fatal to the engine, which makes it a worthless test (other than if the gas lights, and you have been riding with that oil, the engine is likely already toast). This means we don't really have a test for gas in the oil, so if there is any chance of it, it should be changed right away, and don't bother with the test, cause it's worthless.

[natemoen]

Yes my point of the test was to disprove the stick dipping test and I am inclined to agree with you Brian that any gas is bad.

[skids]

Well, you put in the stick, then smell it, and if you want to light it go ahead!
I trust the smell test. Sounds almost Freudian, doesn't it?

[bikerphil]

You can also see if the oil is thinned as how fast it drips off the stick in conjunction with the smell test. Also, there is more of a risk of gas entering the oil if the bike is parked on the centerstand which pitches it forward a bit.

[b.walker5]

Ok then, good experiment. Colour and viscosity is the test I use. If its black, throw it out. If it's thin, throw it out. If it's black and thin, throw it out. If i'm using copious amounts of liquid gold (petrol) and it's not leaking externally, throw it out (and fix the carbs). Rub good oil and bad oil between two fingers, you'll soon get to know the different consistancies. (better done hot, but be careful).

Never have tried lighting a stick dipped in oil, but then I can usually tell if somethings up before it gets to that. Listen to your engine. It's like Sparky and the Train, it talks to you if you know it's language.

[DGXSER]

I will add that any time any one has suggested the stick test, I have always seen it said that if you smell gas, change it!! The stick test has always been an if you think it is there, but can not smell it, try this test.

IMHO, oil is still cheap enough that if you have any doubt, toss it out!! 3000cc of oil, or an engine rebuild........easy choice in my book.

[natemoen]

I totally agree! I never liked it, and this is why I did this, I wanted to show that the test was pointless so it would stop being suggested as a diagnostic tool as it seems it is.