Contaminated gasoline - where to test?

My 2006 Saab 9-3 has been running great. A few days ago, I filled the tank up (14.5 gallons in a 16 gallon capacity tank) and over the next 48 hours, made three very short trips, totaling 12 miles. The next morning, the car would crank but not start. A local mechanic told me no fuel was getting to the engine. Car was towed to the dealer.

Mechanics determined the tank was filled with contaminated gas. They flushed the tank out, and replaced fuel pump. Car has been running great ever since.

Now, I KNOW that many people blame “bad gas” or “water in the gas” as the reason their engines have trouble running smoothly. But in this case, the car stopped running entirely. I can see the contaminant in the gasoline. It is a liquid and is completely immiscible with the gasoline. If you shake it up, the contaminant quickly settles out to a separate liquid layer below the gasoline.

I also know that a test tube with the two layers, gasoline and contaminant, does not freeze but both layers remain a liquid at 15 degrees fahrenheit.

Manager for the gas station owner, even before he heard the whole story, vehemently denied it was his gasoline which caused the problem. He was very aggressive and obviously had his mind already made up, but did agree to have a sample tested, which would prove that the dealership mechanics “are full of sh*t.” I gave him a sample but have not heard back.

How can I get this gasoline tested and identify the contaminant? The Pennsylvania EPA, consumer affairs, and bureau of weights and measures are all uninterested. (Pennsylvania is one of three states which do not regulate gasoline quality. You can legally sell 83 octane gas and call it 91 octane if you want to.) I’ve spent hours calling analytical labs, but no one says they have the capability to test it. One lab even said that since it was in gasoline, and they didn’t know what it was, they couldn’t test it because it could potentially damage their equipment. An analytical chemistry professor at a local university told me that either a GC mass spectrometer, or Nuclear Magnetic Resonance could be used, but they, being a non profit, couldn’t run samples from the public.

Does anyone have any experience with contaminated gasoline? How can I get the substance identified? How can I get the owner of the station, or anyone, for that matter, to take note?

If it is a chain or major brand work your way up the ladder, we had an episode where numerous cars failed after filling with what the carrier determined was bad gas and paid for repairs. It was a BP if I recall correctly. If they get many voices they may indeed determine there was a problem.

I know ths is just a longshot wild guess, but perhaps a gasoline distribution center could refer you to an independent lab that could test this.

A “bad gas” diagnosis always bugs the crap out of me. Very seldom is the case and when a sample from the tank is not preserved for the car owner this makes the diagnosis even more suspect.

If the gas was bad enough to reguire flushing of the tank and fuel pump replacement this means that it should have been running badly on the 3 trips before the car quit.

Maybe what happened is that there a problem in the fuel pump controls and after the car was towed to the dealer it just happened to start. Needing a scapegoat, the bad gas diagnosis may have come into play and the pump replaced as a “possible”.

I’ve had some experience with contaminated gasoline as a tech and the story doesn’t quite wash with me. At this point I side with the gas station owner until proven otherwise.
As I said, any shop or mechanic that claims bad gas better be prepared to back it up with a 1 quart sample. If gas is bad enough to do what they said it did then it will be very obvious to the naked eye and by the sense of smell. JMHO anyway.

Your car was fixed and there is no possible way to determine if a gas station is responsible for a guess from a mechanic. You won’t catch a rainbow and display it in an empty jar either. Some of the other posters have some worthwhile comments.

Has there been a proper “chain of evidience” in use here? Were you present when the sample was removed from the tank? (maybe you were,I don’t know).

What I always ask in “bad gas” cases is, are there other cars affected? It is unlikely just one car was affected.It would sure help your case if a group of people all claim such a condition.

If the gas was bad enough to reguire flushing of the tank and fuel pump replacement this means that it should have been running badly on the 3 trips before the car quit.

The car was running just fine. Then, I put in 14.5 gallons into a 16 gallon tank. Twelve miles later, I have a serious problem. The car probably ran fine on the three prior trips, 4 miles each, because of fuel already in the line. Also, the night before the car did not start, the temperature dropped to 5 degrees fahrenheit. I suspect this contaminant, at that low temperature, gelled and clogged the fuel pump. That’s only a theory. If I can find out what the contaminant is, I can learn its chemical properties, and then decide whether that theory is plausible.

No, I wasn’t there when they pulled the gas out of the tank. But if you are suggesting the mechanics intentionally contaminated the gas as an excuse, why would they do that? There is no motive to do so. They fixed the car by replacing the filter and flushing out the gas tank. It runs great now.

It IS OBVIOUS to the naked eye that something is very wrong with this gas. And I have a 1 quart sample.

As I tried to point out in my original post, I know “bad gas” is overused as an excuse. But that doesn’t mean it doesn’t happen. And in this case, I have more possible proof than the other 99% of claims you hear about.

“There is a principle which is a bar against all information, which is a proof against all argument, and which cannot fail to keep a man in everlasting ignorance – that principle is condemnation before investigation.” (Herbert Spencer)

I’m trying to investigate this. You’ve got experience as a tech. How do I do it?

Were other cars affected? Who knows? Unless they filled up 90% or more of the tank with gasoline that day, and then had that amount in the tank when the temperature dropped to 5 degrees, maybe it didn’t cause a complete shutdown. If they did, how do I find out about those cases?

No, I’m not suggesting the shop intentionally contaminated the sample at all. I’m only pointing out that if the gasoline was bad enough to require a tank drain and flush along with a filter replacement (you previously stated pump, not filter) then the car should have been exhibiting symptoms shortly after you filled it up; as in a few minutes.

Your car would not even have started to make it 12 full miles on the fuel in the lines. Just an FYI here, but the fuel system has a return line to the tank and the fuel that is pumped from the tank to the engine makes a round trip back to the tank.
This is occurring constantly while the engine is running.
Any “good” gasoline in the lines would be purged back to the tank and replaced by the alleged “bad” gasoline in a minute.

Now that you’ve stated you have a sample, what does this sample look like?
Does it have a normal gasoline smell?
If you suspect this gasoline is not up to par so to speak, then try this.
Go out on the driveway and pour a tablespoon or so on the pavement. Quickly cap and set the jar out of the way. Throw a match on the small puddle of gasoline.
If it goes up in a flame quickly the gas is fine; if it is slow to ignite or appears to burn slowly then there could be something wrong with it.

The part I’m having a lot of trouble with here is reconciling gas so bad that it kills a car but yet you managed 12 miles and 3 trouble free trips on it.

The problem with having a “sample”, is that it’s not a true sample. If you sample a piece of cake, and your “sample” is just icing on your finger, is that a true sample of that cake?
Any gas tank will have water at the bottom of the tank. If you take a “sample” from the bottom of the tank (pretend that there was a bottom drain plug) that won’t be a representative sample, will it?
All of the gas (fuel) in the tank, to be a true sample, needs the James Bond cocktail treatment: “Stirred, not shaken” until it’s thoroughly mixed.
So, you see some of the problems with obtaining a sample.

I like to “brainstorm”(you know the technique taught in Business and Management classes) where all idead are initally allowed then the absurd are eliminated (one persons absurd idea may stimulate a good idea from someone else). My “brainstorm” idea is ask the owner of the station if you can post a flyer asking if others had a problem,another idea is you could picket his station,make a sandwich board for yourself with some catchy phrase on it.

Brainstorming about the possible contaminent leaves me empty,do you think it is likely the contaminent came from the delivery truck? was it present by someones planed action or by accident? what product would someone choose if they wanted to do damage?

I’m looking for a qualitative analysis, not quantitative. What is this stuff? I’d like to know. It sure as heck shouldn’t be there.

ANY gas tank will have water in it? I thought modern cars had pressured tank with filtered valves. Are there analytical studies showing that most gas tanks have water in them? What percentage by volume? What percentage is needed before the engine performance is affected? Are these studies published? In a refereed journal? Have they been peer reviewed?

the lower layer could be excess ethanol …did you use dry gas before filling up?

I have never used dry gas in this car (or any other fuel additive). 90% of the time I get my gasoline from this one station. The last two fill ups were also at this station. Dry gas contains either methanol or isopropyl alcohol, I believe.

During the last fill up, I had to wait several minutes before the gas started pumping - another customer was waiting for the fuel to start flowing, also. The attendant, who was outside on a smoke break, had to run back inside and punch a few reset buttons in order to get the gasoline flowing. Later when I asked the manager about this, he told me they had been having problems with the tank alarm, and that is why the gas stopped flowing and a reset button had to be activated. He told me this has nothing to do with the gasoline in the storage tank.

The sample is two layers of liquid at temperatures from 15 to 70 degrees Fahrenheit. The top layer is gasoline. The contaminant layer looks clear, is denser than the gasoline, and is volatile - a drop placed on a glass surface evaporates in 30 seconds. It has an odor to it but that could be from having some gasline dissolved in it; but it doesn’t smell like a gasoline sample. When the contaminant is placed in water, it appears to dissolve, but it imparts a slightly turbid, white appearance to the water, without any visible particulate matter or precipitates.

For your questions, you could try the American Petroleum Institute, or its ilk.
If you did a search of the papers of the American Society of Automotive Engineers, or its ilk, you could get the information on the fuel tank contents, spec’ed and unspec’ed.


You are talking about a philosophical concept known as Occam’s Razor.

Car ran great for 25,000 miles. It continues to run great, with only 1.5 gallons of gas in the tank. Then, 14.5 gallons are added. 12 miles later, car shuts down, will not start. Gasoline is removed and found to be visibly contaminated with an insoluble liquid. Fuel pump (pump, not filter) is found to be not working.

You can come up with all sorts of theories as to what happened. But the simplest, and most logical, is that SOMETHING in the last 14.5 gallons of gas prevented the car from running.

Now, there are certainly variables here. The first 48 hours, when the car was driven only 12 miles, the temperature was at freezing or slightly below. The morning the car would not start, (though engine cranked just fine) it was 5 degrees. Is it possible that only at that cold temperature, the contaminant separated out of solution and formed its own layer on the bottom? Yes, possible, but the original theory, something in the last fill up caused this, is still the most likely.

Even though the diagnosis of “bad gas” irritates some, it does happen. I’ve recently spoken to someone who had sand in their gas (it was swept into the tank when the fill cap to the underground tanks was left off and sand was put down in the winter) and to another person who learned that diesel fuel was accidentally delivered into the gasoline tank at his service station.

I might have agreed with you from the get-go if you have provided all of the information in the original post. You had a sample; why didn’t you say so?
You state the sample has 2 layers? Why not say so?

Contaminated with a flammable liquid or not, I still don’t see what that would have to do with an inoperative fuel pump.

About all I could recommend is lodge a complaint with the state agency that regulates gasoline pumps. In OK it’s the OK Tax Commission. No idea who controls that in your state.

Go back and read my very first post - the one which started this thread.

In it, I state that I have a sample, and I also gave a sample to the gas station owner. I also said I can see two separate, liquid layers in the sample. I also said that the contaminant does not freeze at 15 degrees. Its all there in the first post.

I have also figured out the problem. My hypothesis answers all the questions - why I was able to drive 12 miles without problems, and why the car suddenly would not run, and why there are two layers of liquid.

Hint: Although I was purchasing 87 octane gasoline from the pump, what was delivered was not 87 octane gasoline.

Another hint - the contaminant is flammable, but autos cannot run on it, though some have been modified to use it as fuel, even as early as 100 years ago.

The immiscible liquid that’s liquid at 15F is a mixture of water and ethanol.

Water is insoluble in gasoline. Ethanol (anhydrous) is soluble in gasoline AND water, but more soluble in water. Thus, the water contaminant “draws out” the ethanol into the insoluble, lower layer (I’m assuming the offendig layer is on the bottom here).

This is why fuel ethanol has to go to the extra step of “drying” after distillation: water and ethanol form an azeotrope at roughly 96% ethanol, and cannot be further concentrated by distillation. Now, this is plenty strong for use as fuel in absence of petroleum, but mixing 96% with petroleum will cause ALL of the water (and SOME of the ethanol) to settle out.

Here’s the quote from the original post and the way it’s worded I did not read that as actually having a sample; only that “you know…etc, etc.”.

I also know that a test tube with the two layers, gasoline and contaminant, does not freeze but both layers remain a liquid at 15 degrees fahrenheit.

If the bottom layer in the tank is this unknown substance then I see no way the car could have gone 12 miles on what is in the fuel lines and filter. That would be gone in a mile or two and the car should have been running somewhat rough.
I’ve actually seen a couple of cases (VWs) in which an odd substance (NOT Ethanol, no doubt about it) was pumped into the fuel tank during fillup with no determination as to what that stuff ever really was. However, in both cases the cars started running badly within a mile and only one made it to the shop. The other eventually quit and was towed.

This accumulation of water-alcohol is probably the cause of most instances of “bad gas”. The water-alcohol accumulation could occur over many months. The last fuel fill-up (the “straw which broke the camels back”) gets the blame for what, actually, took a long time (weeks, months) to accumulate. This is one reason we need drains (or, drain plugs) on gas tanks.
If we can’t have drain plugs on fuel tanks, it would be helpful if there was a water separator in the fuel line before the fuel/water gets to the engine.