Amount of alcohol in gas

[Mad Professor]

The question of how much ethanol is really in 10% blend came up on another thread.

A couple of sites give a simple method to get an idea how much ethanol is present. The basis for the test is that ethanol in gasoline can be extracted into a separate water phase (aqueous) and the change in volume of the gasoline phase (organic) can be used to calculate the amount of ethanol originally present in the gasoline. The only equipment needed is a graduated cylinder from a chemistry set:

1) http://www.aviationfuel.org/faqs/alcohol_testing.asp

2) http://www.ehow.com/how_2143193_test-fuels-alcohol-content.html

One site claims the accuracy is +/- 1% V:V. It is unclear wether they compared this method with more accurate means of quantification to support this claim.

 

[habanero]

I'm an analytical chemist and that's a reasonable test, but you have to be pretty careful in your measurements to get plus/minus 1% accuracy. I can think of a lot of potential inaccuracies (other water-soluble components giving falsely high results to name one), but for a quick and dirty test it probably would work okay. My specialty is gas chromatography, so I could easily sneak a few samples through to see how accurate this test is.

 

[Mad Professor]

Habanero, I am also a chemist (12th year teaching college).

My thoughts for a comparison of the methods would be to spike a gasoline sample(s) by serial addition and compare results of the extraction(s) with GC/MS analyses. The MS detector will eliminate deviations in detector response between components (e.g. no standards needed). We have plenty of absolute ethanol (100% undenatured) suitable for spiking.

I also wonder if they bother to remove the 5% water from "grain alcohol" when they make the corn fed fuel and in fact have also made fuel with 0.5% water?

Want to "collaborate on a paper" we "publish" here?

P.M. if you are interested and we'll get a plan together

 

[Tim321]

No water in alcohol

Yes they remove all water from the grain alcohol in the US it call neat alcohol. But I wouldn't be surprised if the gasoline had more than .5% water in it. Brazil uses hydrous ethanol. They can get away with it because all gas has 22% alcohol. The higher alcohol content keeps the water form separating when mixed with gas.

 

[Mad Professor]

Yes they remove all water from the grain alcohol in the US it call neat alcohol. But I wouldn't be surprised if the gasoline had more than .5% water in it. Brazil uses hydrous ethanol. They can get away with it because all gas has 22% alcohol. The higher alcohol content keeps the water form separating when mixed with gas.

It (using lots of grain/hydrous) also makes it more prone to pick up moisture from the atmosphere which eventually leads to phase separation of the water (e.g. a "puddle at the bottom of the tank). This is why aircraft do not fly on corn fed fuel.

 

[2000ssm6]

Hmmmm, how much alcohol in my glass???

 

[huskydave]

None


It would be hard for me to buy ethanol gas only one station sells it and thats the way I want to keep it producing ethanol by growing corn wastes too much energy.

 

[Tim321]

None


It would be hard for me to buy ethanol gas only one station sells it and thats the way I want to keep it producing ethanol by growing corn wastes too much energy.

The energy gain is 20% to 35%. It's not a cure all but a small step to better energy sources.

 

[habanero]

Yes they remove all water from the grain alcohol in the US it call neat alcohol. But I wouldn't be surprised if the gasoline had more than .5% water in it. Brazil uses hydrous ethanol. They can get away with it because all gas has 22% alcohol. The higher alcohol content keeps the water form separating when mixed with gas.

Yep, you're correct. I don't know if gasoline without any oxygenated additives would have that much water or not. True gasoline hydrocarbons shouldn't allow much water in suspension. Now once it has an ether or alcohol oxygenated additive, all bets are off.

Professor,
I won't use an MS for quantitation work. First, you have such limited linear dynamic range it usually necessitates more dilutions (which as you know compound errors). Second, the source is so finicky that every day you have to run a full calibration to be sure you aren't getting response changes. Granted FID response is different among compounds, but the response is stable virtually forever. We have analyses we have done each week for the last 15 years (by a total of 3 chemists on 2 different instruments) and the response factor hasn't varied by more than 2% over the course of that time.

Standards are no problem, we're only looking for ethanol. I'm not going to quant anything other than that, as frankly I don't have the time and it isn't necessary. Like you said, we can spike a gasoline sample with various amounts of ethanol (say from 2-20%) and analyze those by the extraction method, then shoot a sample of each on the gc and compare. If we get wildly different results, then it might be good to go back and analyze samples of the aqueous and organic layers of the extraction as well, but I don't even know if that's necessary.

At any rate, if you want to do the spiking and the extraction method, I can certainly handle the GC work.

 

[huskydave]

The energy gain is not that high when you factor in transportation costs and costs to produce the ethanol. I think corn is inefficient as well seems that sugarcane works much better and some research is being done to use grasses instead of corn. Think about fuel to get the farmer all his supplies fuel used to transport the crop fuel used to transport the ethanol and then transport it again to a gasstation. The numbers don't add up.

 

[Tim321]

The energy gain is not that high when you factor in transportation costs and costs to produce the ethanol. I think corn is inefficient as well seems that sugarcane works much better and some research is being done to use grasses instead of corn. Think about fuel to get the farmer all his supplies fuel used to transport the crop fuel used to transport the ethanol and then transport it again to a gasstation. The numbers don't add up.

That gain includes all cost. Even the food the farmer eats is in some studies. They have to include solar energy that the corn uses to come out with a negative energy balance. The efficacy of grow corn has increased dramatically in the last 20 years as well as the ethanol plants them self are becoming more efficient every year.

 

[B_Turner]

I guess it's supply and demand, but these new uses for alcohol have driven up the consumer price of products like denatured alcohol. I use it in my work, and the price of a DA in places like Home Depot these days is around $14 a gallon. Just a few years ago it was less than $5.

I am trying to find somewhere around here to buy DA alcohol in bulk. Right now to fill a 20 gallon container with DA cost me around $300!

 

[greengoblin]

It (using lots of grain/hydrous) also makes it more prone to pick up moisture from the atmosphere which eventually leads to phase separation of the water (e.g. a "puddle at the bottom of the tank). This is why aircraft do not fly on corn fed fuel.

Is this true of biodiesel as well, i have heard rumors that some of the major airlines may go to biodiesel. If true, we would need to have land about the size of florida to accomodate the demand, in addition to what we already have

With the grain prices where they are, it make no sense financially to produce ethanol anymore. A few places around here have shut down construction of new plants due to that. I personally believe big oil is raising the price to stamp out the competition

 

[Mad Professor]

First test with the extractive method of alcohol determination in "gasoline".

I made a mock 10% gasahol sample consisting of 7.0-mL ethanol diluted to 70.0-mL with Skelly C solvent (e.g. mixed hydrocarbons, slightly lower b.p. range than typical gasoline). To this was added 30.0-mL water and the mixture shaken and allowed to settle. The resulting volumes of the water and hydrcarbon layers were 37.0- and 63.0-mL , respectively. This indicates the transfer of ethanol is nearly quantitative to the water phase within the limitations of the volumetric glassware used for measuring (100-mL vol. cylinder).

(63.0-mL/70.0-mL) X 100% = 90.0 % "gasoline", supra 10% ethanol

Where 63.0-mL = extracted hydrocarbon layer volume and 70.0-mL = initial alcohol/hydrocarbon volume

Conclusion: It seems in the absence of other components this is a good way to determine ethyl alcohol content in hydrocarbon mixtures.

 

[habanero]

The only comment I would have is "gasoline" isn't exactly just a hydrocarbon mixture. I don't know what how much the different additive packages would affect the results. Probably not enough to matter much, but I don't know.

 

[Mad Professor]

The only comment I would have is "gasoline" isn't exactly just a hydrocarbon mixture. I don't know what how much the different additive packages would affect the results. Probably not enough to matter much, but I don't know.

Agreed, and they also have a higher content of aromatics (toluene, xylenes, etc). I did not have a non-gasahol (e.g. AV gas) sample to test as baseline or a sample of corn fed for serial addition. All nearby states have went to corn fed fuel............

Next test will be serial addition of ethanol to corn fed fuel and maybe some AV gas for a comparison if I get to the airport.

Concerning the GC/MS we calibrate ours daily as it get lots of use. For a gasoline sample I'd dissolve in a very volatile solvent (i.e. ethyl ether) and set a solvent delay for the detector to kick in just after the solvent tail clears. Set the column temp initially very low isothermal followed by an increasing gradient. Adjustment of the split ratio and injection volume can keep the detector in the linear range.

With these sort of conditions we get nearly identical integration numbers to those of GC using FID detection. We look at a number of accelerants (fuels) this way in our forensics class when dealing with arson cases.

 

[Ray Bennett]

This thread is way over my head but I have a simple solution for the 10% alcohol in gas. I simply sip out the 10% alcohol on the way home then Im left with pure gas. Just keep in mind this can not be done after the 2 stroke oil is added.:hmm3grin2orange: :hmm3grin2orange: :hmm3grin2orange:

 

[Woodie]

My thoughts for a comparison of the methods would be to spike a gasoline sample(s) by serial addition and compare results of the extraction(s) with GC/MS analyses. The MS detector will eliminate deviations in detector response between components (e.g. no standards needed).

I was gonna say the exact same thing!!!

Right up to the part where you got to the, like, science and stuff...

 

[Grande Dog]

The question of how much ethanol is really in 10% blend came up on another thread.

A couple of sites give a simple method to get an idea how much ethanol is present. The basis for the test is that ethanol in gasoline can be extracted into a separate water phase (aqueous) and the change in volume of the gasoline phase (organic) can be used to calculate the amount of ethanol originally present in the gasoline. The only equipment needed is a graduated cylinder from a chemistry set:

1) http://www.aviationfuel.org/faqs/alcohol_testing.asp

2) http://www.ehow.com/how_2143193_test-fuels-alcohol-content.html

One site claims the accuracy is +/- 1% V:V. It is unclear wether they compared this method with more accurate means of quantification to support this claim.

We've got these. http://www.baileysonline.com/itemdetail.asp?item=18454&catID=

 

[habanero]

...Concerning the GC/MS we calibrate ours daily as it get lots of use. For a gasoline sample I'd dissolve in a very volatile solvent (i.e. ethyl ether) and set a solvent delay for the detector to kick in just after the solvent tail clears. Set the column temp initially very low isothermal followed by an increasing gradient. Adjustment of the split ratio and injection volume can keep the detector in the linear range.

With these sort of conditions we get nearly identical integration numbers to those of GC using FID detection. We look at a number of accelerants (fuels) this way in our forensics class when dealing with arson cases.

I didn't say quant work can't be done with the mass spec-at a former job I ran thousands of samples using single ion monitoring to quant a compound found in wheat germ. But day-in, day-out quant work, the FID is faster and much cheaper in the long run. We have both in our lab and unknowns go through the mass spec, then get quanted on the FID.

As for the GC method, with the FID instrument I'd use a high-boiling solvent like DMAC so all the peaks of interest are out before the solvent comes through. Might be a bit of a trick to find a good internal standard, but you should have plenty of room to fit something in.