Beware fuel system icing - it happened to me...

[SpamPilot]

Flying my M20K 231/Rocket 305 through eastern Washington this New Years weekend at 14500 MSL in VFR conditions, I felt some light turbulence.  After about a minute, it occurred to me that the turbulence was not changing in amplitude like turbulence usually does.  I considered whether it could be a small prop imbalance due to icing, but that would be strange since I had been in clear air since takeoff an hour earler.  I considered whether it could be a cylinder misfiring, but EGTs were normal on all cylinders, and TIT remained where I had dialed it in.  Manifold pressure remained at 30 inches, engine speed was still 2400 RPM, and I wasn't losing airspeed.

Then the engine sneezed.  That got my attention.  I aimed towards the nearest airport, pulled out three inches of manifold pressure, pointed the nose down, and popped the speedbrakes.  The engine sneezed one more time on descent, but I stayed focused enough to run all my checklists and grease the landing.

Maintenance wasn't available on the weekend, so the aircraft sat on the ramp, shivering ignominiously in sub-20F temps.  First thing Monday morning, we brought her into the hangar and immediately pulled a belly pan and opened the gascolator/sump to look at the fuel strainer screen.  It was partially coated with what looked like a very thin paste - imagine spreading the thinnest possible layer of white or maybe grey toothpaste over the screen, then wiping it off so you could see through it.  Over the next 30 seconds, in the warmish hangar, the paste melted and turned into beads of water.

This topic has been covered on the forum in the past (e.g. So whats too cold for the plane  and IPA Winter Ops), but I figured a fresh PIREP is always a good PSA.  In reading through those topics, I noticed a couple comments that deserve a response.

kortopates wrote "I really don't believe there is risk of fuel becoming saturated with water unless there is more moisture than just humid air in the tanks" but then, in a very gentlemanly and helpful way, he retracted his statement and shared Mike Busch's observations:

He [MikeB] told me there is a real risk of dissolved (undrainable) water coming out of solution at low temperatures and freezing into ice crystals that can cause engine stoppage. He knows of several emergencies that occurred because of this, although it's definitely not common. Mike says these incidents don't occur in the wintertime as far as he knows. He says the major risk is taking off from a low altitude airport where the temperature and humidity are high and the fuel has become heat-soaked because the aircraft has been sitting in the sun for some hours, then climbing up to the Flight Levels where the OAT is very cold. It takes very hot fuel that can hold a lot more water in a dissolved state (as you pointed out before). The conditions leading to this issue have always been summertime conditions though, not winter. 

I will simply add that 1) it's winter now, and very very cold for the Pacific Northwest, and 2) I fueled (almost 60 gallons to top off) from an above-ground tank that was at below-freezing temperatures.  Thus, I have an existence proof that it can happen in winter with cold fuel.

jlunseth wrote of his experience never having this problem: "Fuel gets put on board at ground temps which are warm, and in relatively humid conditions.  Then I go to altitude, stay there for 4 hours at -25C, and nothing happens. "

Well, this happened to me after only 1 hour at altitude, and yes the OAT at that altitude was -25C (=-13F).  I do have many hundreds of trouble-free hours flying this aircraft and another Mooney at temperatures significantly below freezing, though usually never lower than -10C.  This happened to be the first time for me in this aircraft flying at -25C.

I had done a thorough sump/drain before my flight and there was no evidence of liquid water at all.  After observing the ice crystals in the strainer screen, we drained half a gallon from the wings and the sump and there was still no evidence of water.

For those who haven't followed the threads, the culprit is water dissolved in the fuel, not free liquid water.  100LL can't dissolve much water, but water is not completely insoluble in fuel, and having tiny amounts of water (on the order of 30 parts per million) dissolved in the fuel is OK so long as it remains dissolved.  Sub-freezing temperatures and nucleation sites (perhaps even localized turbulence as fuel flows through a fine mesh) will cause the water to precipitate out as microscopic ice crystals.  Once an ice crystal forms, it's going to continue to grow as more fuel flows past it and dissolved water preferentially attaches itself to the crystal structure.  These crystals can attach themselves to rough surfaces, get caught in fine mesh screens, or plug tiny injector orifices.

For you chemists who are concerned about the proper use of the word "dissolved", I admit I don't know if the water is truly dissolved in the gasoline (= in solution on a molecular scale) or simply finely divided and in suspension (= extremely tiny water droplets that are still orders of magnitude bigger than molecular scale).  AC 20-125 discusses "entrained" water and describes it as the latter.  I will note that our white bucket test revealed the gasoline in my tanks to be both "clear=clean" and "bright=dry", with no cloudiness that would indicate entrained water.

The solutions to the problem, as have been discussed elsewhere on this forum, are:

1) If the fuel is dry, ice can't form.  Read AC 20-125 to understand the various ways water can get introduced into fuel.  Use fuel from trustworthy tanking/pumping systems that are properly maintained.  Make sure your gas cap O-rings are in good shape.  Always sump your fuel to check for water.
2) If the fuel has water but remains above freezing as it travels through the aircraft fuel system, then small amounts (30 ppm or less) probably won't have any effect on operation.
3) Adding anhydrous isopropyl alcohol (IPA) to the fuel, in amounts not to exceed Mooney SI M20-64, encourages the water to dissolve into the IPA, which lowers the freezing point of the water, hopefully enough to prevent it from precipitating out as ice.
4) (my recommendation) After taking off from hot and humid conditions, or even cold and dry conditions, when flying for as little as M minutes in temperatures below N degrees C, choose an altitude and a route that give you good descent and landing options in case, like me, you encounter fuel system icing despite all the normal precautions.  In my case, M was 60 and N was -25.  I am congratulating myself on electing to perform that flight on that day only because it was VMC and, in the event of a problem, I wouldn't have to descend through several thousand feet of icing conditions.  I am also considering myself lucky that it didn't happen 30 minutes later over the Cascade mountains.

Edited January 4, 2022 by SpamPilot

[Fly Boomer]

Great post, and great outcome.  I don't think it's relevant to your experience, but I'm curious what year your Rocket was built (by Mooney, not Rocket Engineering).

[ArtVandelay]

5) Live and fly in warmer climates. ;-)

Glad it worked out well, I always assume if no water found when sampling the tanks there’s no water. Not necessarily true I guess.

[PT20J]

When I worked at a flight school in Anchorage, it was a company SOP to add isopropyl alcohol to the tanks after each refueling in the winter.

From the M20J Service and Maintenance Manual. 

[mhrivnak]

10 minutes ago, PT20J said:

When I worked at a flight school in Anchorage, it was a company SOP to add isopropyl alcohol to the tanks after each refueling in the winter.

From the M20J Service and Maintenance Manual. 

I see that 1% by volume is the max, but what's a normal ratio to add?

[jaylw314]

@SpamPilot, you mentioned you got fuel from an above ground tank in freezing conditions.  If the supply tank ALREADY had microscopic ice crystals in it, sumping the fuel wouldn't have helped a dang bit, and once you warmed up the fuel, if the water could dissolve in the fuel, there'd be nothing to sump once it warmed up.

If so, it's possible others who used that fuel might be vulnerable in similar conditions.  Do you think this would be worthy of an ASRS report or even contacting your local FSDO?

[PT20J]

18 minutes ago, mhrivnak said:

I see that 1% by volume is the max, but what's a normal ratio to add?

Good question. It's been a long time and I don't recall what Aero Tech specified in the SOP. I might go with half a percent just to avoid exceeding the max and assuming that the amount of water is small.

[EricJ]

25 minutes ago, PT20J said:

Good question. It's been a long time and I don't recall what Aero Tech specified in the SOP. I might go with half a percent just to avoid exceeding the max and assuming that the amount of water is small.

The Mooney SI says 1% for anything with a Lycoming, since that's the limitation in Lycoming's guidance, but has stated up to 3% for K models.   So the1% limit is for the engine, perhaps gasket integrity or something.

 

[ilovecornfields]

Maybe one of the A&Ps can comment (like @kortopates) but would pulling the gascolator drain in-flight get rid of those ice crystals?

[PT20J]

17 hours ago, ilovecornfields said:

Maybe one of the A&Ps can comment (like @kortopates) but would pulling the gascolator drain in-flight get rid of those ice crystals?

The drain only empties the sediment bowl which collects liquid water and particulates. During operation, fuel from the tanks fills the bowl and exits the top through a filter screen which is what collected the ice crystals. So, there is no flushing action through the screen by opening the drain.

Skip

[A64Pilot]

Different fuel but in the Army before any fuel could be used at the beginning of the day, an Aqua Glo test was done, it determined the parts per million of dissolved water and if above allowed limits then the truck or fuel point would have to recirculate the fuel through the filter until it was within limits.

it wasn’t all that uncommon to have to recirculate fuel.

https://www.gammontechstore.com/SearchResults.asp?Cat=1896

The filter apparently contained the same or similar substance to a baby diaper, crystals that turned into a gel when they absorbed water.

I’ve never seen or heard of this being done in the Civilian world and have wondered about it.

But yes, water is miscible I think the word is in fuel, any petroleum fuel, particularly any fuel containing alcohol. so while alcohol can and will cause any free water to dissolve in the fuel. if left in the fuel over time it can attract water.

Edited January 5, 2022 by A64Pilot

[jaylw314]

1 hour ago, A64Pilot said:

Different fuel but in the Army before any fuel could be used at the beginning of the day, an Aqua Glo test was done, it determined the parts per million of dissolved water and if above allowed limits then the truck or fuel point would have to recirculate the fuel through the filter until it was within limits.

it wasn’t all that uncommon to have to recirculate fuel.

https://www.gammontechstore.com/SearchResults.asp?Cat=1896

The filter apparently contained the same or similar substance to a baby diaper, crystals that turned into a gel when they absorbed water.

I’ve never seen or heard of this being done in the Civilian world and have wondered about it.

But yes, water is miscible I think the word is in fuel, any petroleum fuel, particularly any fuel containing alcohol. so while alcohol can and will cause any free water to dissolve in the fuel. if left in the fuel over time it can attract water.

The super-absorbent polymer they use for jet fuel is great for absorbing water, but I recall there was an airline accident where the polymer broke out of a filter and contaminated the fuel.

FWIW, miscibility is the same as dissolving, it's a truly homogenous solution.  Water is barely miscible in gasoline (about a tablespoon in a tank's worth of avgas), so hard to imagine it could have been the culprit.  Suspended or entrained water can be a lot more, though numbers are hard to come by and probably vary with a lot of factors.

https://www.faa.gov/documentLibrary/media/Advisory_Circular/AC20-125.pdf

Edited January 5, 2022 by jaylw314

[A64Pilot]

Add alcohol to Avgas or Mogas and it’s ability to absorb water increases, by quite a bit.

 That is the purpose of adding the alcohol as a de-icer, it mixes with the water and fuel ,makes water miscible in fuel and the mix burns with no problem, and of course won’t freeze.

 But, and there is always a but, if the alcohol is mixed well in advance, it can absorb water from the atmosphere and in low pressure and or low temps the water can come out of suspension, although not aviation related this seems to explain it well, look at how much more water can be absorbed with just a little alcohol.

https://petroclear.com/resources/dont-be-phased.php

So if you use it as a de-icer, burn the fuel in a reasonable time and don’t use it for storage.

[aviatoreb]

On 1/4/2022 at 2:49 PM, SpamPilot said:

Flying my M20K 231/Rocket 305 through eastern Washington this New Years weekend at 14500 MSL in VFR conditions, I felt some light turbulence.  ...

Great post SpamPilot.  Great job handling the situation.  You had a rather unusual circumstance of actually definitely finding the ice and watching it melt away.  I presume in many situations the evidence melts away before it can be seen and then one is left wondering!

Right - its the temps at altitude that are important.

I have been using IPA in the fuel in recommended ratios for years.  I have not had this particular problem myself, but I have known others to have had it and that is good enough for me to include IPA in my operation.

By the way - I like these - since they are pre-measured and right sized for one tank per sized container.  I don't want to open a large jug of highly pure IPA just to have it gradually have the jug start to retain water before I end up using it.  These are sealed, right sized and easy to travel - add some when I am away.  And we all need to be sure we get the high quality IPA rather than that 70% stuff from the drug store.

https://www.amazon.com/dp/B07J62C1K5?psc=1&ref=ppx_yo2_dt_b_product_details

Edited January 5, 2022 by aviatoreb

[jaylw314]

How do you actually get the IPA mixed with fuel already in the tank?  Obviously, it's easiest if you have a nearly empty tank, you put the IPA in first and then the fuel, but what if the tanks already full or nearly full?  Jump up and down for a while?

[alextstone]

I really appreciate this thread.  I was unaware of the option to add IPA.  I'm headed to Utah to go skiing later this month from MS., In all likelihood, I will depart from warm weather to very cold up high.  I will add IPA now.  Thank you so much for improving the safety of my flying....

 

Alex

[aviatoreb]

1 hour ago, jaylw314 said:

How do you actually get the IPA mixed with fuel already in the tank?  Obviously, it's easiest if you have a nearly empty tank, you put the IPA in first and then the fuel, but what if the tanks already full or nearly full?  Jump up and down for a while?

I do as follows - whether it is an empty tank or a half full tank - I use those little plastic sealed containers I posted as above.  One per side.  If it is added to a full (mostly) empty tank or to a half full tank, its all ok, its within recommended range.  Don't just dump it in and then start fueling or vice versa don't fuel and then dump it in.  I think its likely much better mixed in if you pour in gradually with one hand while operating the fuel pump with the other hand pouring IPA gradually into the fuel stream.  That way at least it is well mixed into some of the fuel and then I presume as you fly around it mixes further as the wings roll back and forth.  And diffusion over time if sitting in your hangar.  Yah - imperfect but better than dumping it in.

[aviatoreb]

28 minutes ago, alextstone said:

I really appreciate this thread.  I was unaware of the option to add IPA.  I'm headed to Utah to go skiing later this month from MS., In all likelihood, I will depart from warm weather to very cold up high.  I will add IPA now.  Thank you so much for improving the safety of my flying....

 

Alex

I'm just riffing here on what I believe might be true and how I operate so feel free to pick it apart,

I feel if the wing is warm soaked then it would take more than a few minutes to freeze a tiny amount of water in the fuel since it would take more than a few minutes to lower all that fuel in your wing to the freezing point of the water - and oh by the way - what is the freezing point of water dissolved in fuel?  I bet it is a fair bit lower than 32F?  So I just don't worry about it on a shortish flight in the summer.

In the winter, the wing is already very cold and going to get even colder.  So I use it all the time at least 6 months of the year here in this part of the east they call "The North Country".

[Browncbr1]

Seems like I read someone on MS using HEET??

[aviatoreb]

4 minutes ago, Browncbr1 said:

Seems like I read someone on MS using HEET??

Straight IPA is actually an FAA approved substance and it is in the continental poh.  So I wouldn't go with heet not knowing exactly everything that is in there, it being more expensive and also no need since high quality IPA is cheaper and sufficient anyway.

[Browncbr1]

7 minutes ago, aviatoreb said:

Straight IPA is actually an FAA approved substance and it is in the continental poh.  So I wouldn't go with heet not knowing exactly everything that is in there, it being more expensive and also no need since high quality IPA is cheaper and sufficient anyway.

I agree,  I just thought I saw someone doing that to remedy hiccups that were attributed to undissolved water drops that get dislodged from tanks during turbulence.

[Fly Boomer]

Wonder what the other 1% is in those 99% bottles?

[jaylw314]

41 minutes ago, aviatoreb said:

I do as follows - whether it is an empty tank or a half full tank - I use those little plastic sealed containers I posted as above.  One per side.  If it is added to a full (mostly) empty tank or to a half full tank, its all ok, its within recommended range.  Don't just dump it in and then start fueling or vice versa don't fuel and then dump it in.  I think its likely much better mixed in if you pour in gradually with one hand while operating the fuel pump with the other hand pouring IPA gradually into the fuel stream.  That way at least it is well mixed into some of the fuel and then I presume as you fly around it mixes further as the wings roll back and forth.  And diffusion over time if sitting in your hangar.  Yah - imperfect but better than dumping it in.

I recall from chemistry lab that pouring large volume into small volume mixes better than pouring them simultaneously, but admittedly, that was a lab study, not a fuel tank.  Still, if I have an empty tank, I'd probably try the IPA first, then fuel.  But it's good to know the expectation is that you can still add it to a full tank.

10 minutes ago, Fly Boomer said:

Wonder what the other 1% is in those 99% bottles?

Water.  IPA sucks up water until its about 91%.  You can't even distill it higher than 91%, if you do both the IPA and water boil out at the same rate.  The only way is to run it over a desiccant, so in practice it's really hard to get it to 100%.

FWIW ethanol settles at 95%, which is why everclear is 190 proof and not 200 proof.

Edited January 5, 2022 by jaylw314

[Will.iam]

1 hour ago, aviatoreb said:

I do as follows - whether it is an empty tank or a half full tank - I use those little plastic sealed containers I posted as above.  One per side.  If it is added to a full (mostly) empty tank or to a half full tank, its all ok, its within recommended range.  Don't just dump it in and then start fueling or vice versa don't fuel and then dump it in.  I think its likely much better mixed in if you pour in gradually with one hand while operating the fuel pump with the other hand pouring IPA gradually into the fuel stream.  That way at least it is well mixed into some of the fuel and then I presume as you fly around it mixes further as the wings roll back and forth.  And diffusion over time if sitting in your hangar.  Yah - imperfect but better than dumping it in.

One per side? I guess I’m missing something in my math. One bottle 16oz one gallon 128oz so one bottle only treats 12.5 gallons at 1% so wouldn’t you need at least 2 bottles + per wing? I hold 37.8 in one wing so i need at least 3 bottles for 1%. And my K poh says up to 3% so that would be 9 bottles. So one case wouldn’t even do a full fill up for me. This is getting expensive. 

[jaylw314]

7 minutes ago, Will.iam said:

One per side? I guess I’m missing something in my math. One bottle 16oz one gallon 128oz so one bottle only treats 12.5 gallons at 1% so wouldn’t you need at least 2 bottles + per wing? I hold 37.8 in one wing so i need at least 3 bottles for 1%. And my K poh says up to 3% so that would be 9 bottles. So one case wouldn’t even do a full fill up for me. This is getting expensive. 

You can use less than the maximum allowable.  16 oz of IPA should absorb 4 oz of water, which is a lot of water in a tank....