Biodiesel and SVO Forums Home
simple test for water in fuel.

This topic can be found at:

April 23, 2009, 01:03 PM
simple test for water in fuel.
Living in a damp climate like Ireland makes it difficult to dry biodiesel. I use hot air blown over a bubble shaped fountain to dry the bio to sparkling clear but if I allow moist air back into the tank the bio will become cloudy within a week or so. Problems with water content in dino diesel are not uncommon here at this time of year and can cause serious problems with injection systems.
I have looked at professional moisture testers but they are very expensive so I have attemped to devise an accurate method of measuring water content.
I purchased a plastic drinks cup with a sealable screw cap. I bored a small hole in the lid and fitted three yards of 3 mm pvc hose to it, sealing the joint with a hot glue gun.
Using small cable clips I fixed the pvc tube onto a white painted board in the shape of a U about 3 foot tall.
I half filled the U with water coloured with food colouring.
This was to be my pressure guage.

I put the beaker on a digital scales and measured 250 grams of dried biodiesel into it. I filled a drinks bottle top with calcium carbide powder and carefully floated it on the surface of the bio like a little boat. I screwed the cap back onto the cup without disturbing the calcium carbide and marked the level of the red water in the U tube.

Ihe idea is that when the calcium carbide comes in contact with the water in the bio it will produce acetylene gas which will pressurise the inside of the cup forcing the coloured water to rise up the U tube. This would give me an accurate repeatable reading of water content.
I had guessed that I might get a reading of a couple of inches which would be enough for an accurate reading but I was wrong. Within a minute of mixing the the contents of the cup the red water had risen a full three feet and started to escape out the top of the tube.

So the experiment worked but I have to rethink the scale. as I see it I have three options
A... Weigh out a smaller amount of bio into the cup.
B... Make a pressure guage with the same tubing but about 10 feet tall
C... Make another pressure guage, same height but with larger bore tube.

I think option C makes most sense.

Calcium carbide is easily available on Ebay. The stuff I got was in large lumps but it was fairly easy to pound into powder with a hammer.

I will post a drawing of the setup as Im not sure Ive described it very well.

When I started to shake the contents of the cup the pressure guage began to rise rapidly
April 23, 2009, 01:31 PM
opition C will not work. No matter how large the bore is the Pressure would be the same, Only the volume would change. But then the larger volume of the tube would cause slightly less pressure.

The best would be to decrease the amount of reaction, (less Bio and/or less calcium carbide).

Also something that might be able to use that is easier to get than carbide is Alka=Seltzer. When it contacts water it produces CO2 gas. But not sure how it would react with Bio.


If you can't dazzel them with brilliance, then baffel them with bullchit.
April 23, 2009, 01:48 PM
You are right of course, Its been a long time since I studied physics at school.
I can easily reduce the amount of biodiesel but reducing the amount of calcium carbide is not an option as I have to have more than is required to react with all of the water.
I shall remake the existing pressure guage to about 4 feet tall, failing that Ill have to buy more tube.

Here is the picture, the scale isnt correct but shows the layout.

April 23, 2009, 02:00 PM
Im not clear on something, will the calcium cloride come in contact with the bd or will the water vapor leave the bd and contact the cc. It sounds like it just floats on top in the cap and never touches the bd.
April 23, 2009, 02:16 PM
Raften ..

I gathered it would float till he sealed the container and then shook it up.


If you can't dazzel them with brilliance, then baffel them with bullchit.
April 23, 2009, 03:10 PM
Yes the cap full of calcium carbide floats on the surface of the biodiesel until the screw cap is securely on the cup then you tip it up and mix the contents. This prevents loss of gas while you are screwing on the cap.
I am aware of the other methods of testing for water in fuel but I want to devise a method that can be calibrated to give a numerical result rather than a subjective judgement. The Sandy Brae tester is very good but expensive.
The idea of using Alka Seltzer sound like it could work. I will buy some tomorrow and try it.
April 23, 2009, 04:27 PM

You said you was going to increase the length of the tube to 4feet. I dont think this will be enough. A 48 inch waterlock you only give you max theroetical movement of 24 inches. Because of the curved bottom you will not get the full 24inches. Lets say you can get 22 inches, since 1 inch of water at 39C is .0361 PSI the 22inches would be 0.7489 PSI. I am thinking you would run over this amount of pressure from reading the Sandy Brea instructsions.


If you can't dazzel them with brilliance, then baffel them with bullchit.
April 23, 2009, 05:22 PM
I see that I may need a much longer tube to measure the pressure created by the gas produced. Alka seltzer may produce smaller volume.
I have not attemped to calibrate the system because I have not suceeded in getting a valid reading yet. Once I adjust the gas volume/ pressure guage height to the point where I can get a reading I would propose to start by measuring a sample of dinodiesel which should be at or below 500 ppm. ( ASTM or Din 14014 standard.) I would then adjust the sample by adding 1% water. with these two readings I should be able to calibrate the scale on the pressure guage.
April 23, 2009, 07:18 PM
Jon Heron
Cool idea!
What about just using a standard low pressure gauge like what is used for ATV tire pressure? See here.

Here is a calculator and some more information on manometers. HERE and HERE

I look forward to your results!


Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
April 23, 2009, 07:20 PM
Imake. Cool idea! How about using a pressure gauge like the Sandy Brae? You may be able to use magnets to flip the boat so you dont have to turn it upside down. Is Calcium Carbide what's used in Carbide lamps? What does Sandy Brae use as the active ingredient?
April 23, 2009, 08:15 PM
Calcium carbide is the stuff for makeing actylene in carbide miners lamps. Or long ago used for cutting steel. My dad was an 'oldtime' welder and had an actylene generator, He hauled out 50lb barrels all the time from plant he worked for till he finally decided to buy a couple pressure bottles.

The SandyBrea tester uses Calcium Hydride. Here is the MSDS http://www.utahbiodieselsupply...agent_A_05_15_06.pdf


If you can't dazzel them with brilliance, then baffel them with bullchit.
April 23, 2009, 10:04 PM
Would the device that Imake has come up with be called a manometer or something that sounds like that?
April 24, 2009, 12:15 AM
Imakebiodiesel: one approach that would reduce the gas volume produced to a manageable level is to meter the biodiesel into the Carbide until a predetermined pressure registers on the manometer. A calibrated burrette filled with your biodiesel sample, a 2-hole cork stopper, a beaker with the Calcium Carbide, and your manometer are all that is needed.

The more biodiesel needed to reach a given pressure/displacement, the drier it is. If you dribble in 10cc of biodiesel, and only see 10 cc displacement (after compensating for compression, thermal effects and gas laws), then the fuel is completely dry.

Second idea: loss in weight variant - put the beaker with the calcium carbide on a sensitive scale, along with the sample container of biodiesel. Tare the scale or record the weight. Dump the biodiesel into the calcium carbide and wait for the gas to stop evolving. The escaping gas removes mass that is proportional to the water content (acetylene is C2H2, isn't it?). Someone smart in chemistry can tell us what the change in mass would be for a given amount of water.


ps; I just dumped some biodiesel into some calcium carbide and tiny bubbles are drifting up from the rocks. I see the obvious need to crush it first. This is a brilliant idea. I'll play with it some more.
April 24, 2009, 01:41 AM

I think you hit it with metering in the biodiesel to get a predetermined pressure. That way you meter it in until you have converted a specific mass of water to c2h2. Then it's a simple division to get ppm.

April 24, 2009, 06:03 AM
The device is indeed a manometer, a simple but effective form of pressure guage. I could of course use a dial pressure guage of the sort used for low pressure tyres but Im not sure they would be accurate enough.
The idea of metering the biodiesel into the carbide is interesting but I cant see how to get liquid in without letting gas out.
The grade of carbide is not important as long as you use more carbide than is necessary to react with all of the water present in the fuel. The most important thing is to grind the carbide into a fine dust so that it reacts quickly. A couple of steel ball bearings in the cup will help to mix the contents.
Im not plotting the overthrow of the h-w-h method, Im just experimenting with an alternative which might be convenient and quick.
April 24, 2009, 08:34 AM
I repeated the procedure today but used only 50 gms of biodiesel. After about 3 minutes of gentle agitation I got a reading of 303mm from the manometer, much more satisfactory. Im going to repeat the experiment a couple of times over the weekend to make sure Im getting a consistent result and then try a sample of dry dinodiesel.
When screwing on the lid I get a slight reading as the air inside the cup compresses. Its is important to to zero the scale at that point to avoid getting a false reading. Also avoid warming the cup with the heat from your hands, hold the cap by your fingertips and gently swirl the contents until the manometer stops rising.
April 24, 2009, 08:57 AM
This is a great idea. Whether calcium carbide, alka seltzer, or some other chemical is the final choice the concept is good. Personally I like the idea of weighing the results rather than reading pressure since we all have scales already and pressurizing a combination of biodiesel and acetylene makes me somewhat uncomfortable. Wouldn't want anyone to mix too much by mistake.

Anybody know what the reaction of these chemicals with methanol is? I ask because with drywashing I feel there is some methanol passing through and the traditional measures for water also detects methanol evaporation, and the measures for methanol (w-h-w) also detect water. It would be nice to measure one without the influence of the other.

April 24, 2009, 09:18 AM
Originally posted by HughT:
. . . I like the idea of weighing the results rather than reading pressure since we all have scales already and pressurizing a combination of biodiesel and acetylene makes me somewhat uncomfortable. Wouldn't want anyone to mix too much by mistake.

. . .the traditional measures for water also detects methanol evaporation, and the measures for methanol (w-h-w) also detect water. It would be nice to measure one without the influence of the other.


A simple U tube manometer can be very accurate. I don't think excess pressure is a problem. If the pressure is too great it will just blow the liquid out the end of the tube, allowing the gas to escape to the atmosphere.

Water and methanol are miscible. If you have water you will also have some methanol. If all of your testing is done on biodiesel samples, then the methanol contribution will always be present. But, since it is present in all samples, then it will negate itself so the effect of methanol in the reaction will be, effectivly, zero.

Along a similar vein, when titrating with our stock solutions we don't really care what the actual concentration is as long as we use the same solution in the reaction.
April 24, 2009, 09:21 AM
As far as I know methanol does not react with calcium carbide but perhaps someone with more knowledge of chemistry can confirm that. So this test should be able to measure water only.

Most biodiesel brewers have digital kitchen scales accurate to about one gram. While this is adequate for weighing out 50 grams its not able to detect very small losses of water vapour or gas.

The procedure I have suggested does produce a small quantity of acetylene gas which is highly inflammable but I would think it would not pose a high risk for people already used to handling methanol etc. (no smoking, no naked flames, good ventilation.)
April 24, 2009, 10:03 AM
I repeated the test with the same sample of biodiesel and got a reading of 310mm. This is within 2% of the first result. My digital scales are only accurate to 1 gram, which is 2% of the sample weight so Im pretty happy with that. I want to repeat it again before moving on to dinodiesel etc. However Im going away for the weekend so we will have to wait till Monday before I can start calibrating the manometer.
It would be great if someone would duplicate these tests just to make sure Im not doing something wrong.