Effects of Water in Fuel

by Brian Hampton


Recently there was a question asked in one of the modeling boards on the internet concerning the affect of water in fuel and how much was allowable. As usual, opinions varied enormously but no facts were given to back up them up so I decided to find out for myself. The method was to be simple…I’d use my trusty old SC 1.08 and add predetermined amounts of water to the tank before each flight. Of course, I’d give it a good check on the ground before actually flying.  The fuel tank checked out to be 250cc and I had a graduated plastic syringe which I’d use to inject water when the tank was half full then complete the filling to get it all nicely mixed before starting the engine.

The first flight was done with no water, just my usual 80/20 all castor, to make sure all was well.
For the first attempt at adding water (and not really knowing what % was acceptable) I injected 5cc which gave me almost exactly 2% water mix. The engine started first flick (as usual), idled the same, transition and full throttle were the same so I flew it. I couldn’t detect any difference.

The second attempt (after draining the tank) was to add 7.5cc of water for a 3% mixture. Once more there was no discernible difference so I flew it again.

The third attempt was with 10cc for a 4% mixture. This time the engine was almost impossible to start and even then only by priming with fresh fuel. Once the prime ran out it stopped. Having reached a limit I drained half the tank into a squeeze bottle and topped up with fresh fuel to bring it back to 2% again. Once more the engine ran perfectly.

But the 4% fuel I’d drained into the squeeze bottle was a horrible murky colour, no one would have used it if they’d seen it going into their tank. Some time later I happened to look at the bottle again and this time the fuel was clear but there was a thick layer of oil in the bottom. It was obvious that the discolouration was from tiny droplets of oil which slowly settled to the bottom so I added some fresh fuel to make it close to 3%, shook it up and it remained nice and clear. This indicated to me that if the fuel remains clear when shaken then it was OK to fly with.

But now the plot thickens…..
Back at home I’d put that squeeze bottle on a table and in the evening I noticed that oil had seperated out again. Shake it and it went murky just like the 4% did. I guessed that it was because the evening was cooler so I put the bottle under some hot water for a moment and sure enough..it went clear. OK, now it was obvious that the allowable water was temperature dependent but to what extent?

This time I decided to experiment with the only synthetic fuel I’d ever used….18.5% Mobil Jet Oil 2. I filled the syringe with 30cc of fresh fuel and drew water into the syringe through a plastic tube that held 1cc for every 12cms (in other words, 3cms gave me .075cc which was .25% of the fuel volume). What I was looking for was the lowest temperature at which the fuel remained clear when shaken for a given % of water. I used a thermocouple inside the syringe for the fuel temperature. Now for the results….


.25%   -10C (14F)
.5%     -5C (23F)
.75%    0C (32F)
1.0%    5C (41F)
1.25%  11C (52F)
1.5%    16C (61F)
1.75%  22C (72F)
2%       29C (84F)
2.25%  34C (93F)
2.5%    39C (102F)


These temperatures seemed to be quite specific in that even a 1C rise or fall was enough to make an immediate change from cloudy to clear. However, it might be noted that in my tests at the field I had a problem somewhere above 3% water which doesn’t fit the figures above (the temperature that day was around the 30C mark). This indicates to me that castor comes out of the solution at a higher temperature than the particular synthetic I tested. Considering that I didn’t exactly have a laboratory to do these tests in (unless my fridge freezer counts) there seems to be almost a straight line correlation between water content and the temperature that the fuel goes bad. It would be interesting to know what affect, if any, nitro has on a similar fuel but I’ve never used it.

Comments are closed