Frozen avgas

[James McDiarmid]

Engine Out 4/5/24

Hi all, I had an engine failure 3 weeks ago and I thought I’d share it with you.

My wife and I were 2h into a flight from Exeter to Venice in the cruise at FL230 in my 2008 Mooney Acclaim-S. I bought my second Mooney last year - I’d previously owned an ovation 2GX but the allure of flying faster above the weather led me to purchase an Acclaim. During the flight the weather had been pretty good with the exception of the forecast tailwinds in fact being 50-60kt headwinds and consequently slowing our progress. We were in Belgian airspace and I was talking to Brussels control.

OAT was -32C as we surfed through smooth air in the tops of a thin layer of cloud into a 55kt headwind.

As usuaI was flying by the numbers - at best economy…

We were cruising 28.0” MAP 2300 RPM LOP @15.0gph

In retrospect the early indication of an issue was very occasional rough running which started 20 minutes or so before the engine stopped.

Also - oddly EGT/mixture was not ‘stable’ (the mixture seemed to automatically be getting leaner - EGT increasing above 50 degrees LOP - so I was watching it and enriching in). I had never experienced this before.

As a consequence of the LOP mixture instability I decided to run ROP which provided a little more stable EGT-wise until after around 10 mins the engine stopped abruptly.

At this moment I received the low manifold pressure yellow (?red) warning on G1000.

All 6 EGT/CHT immediately dropped to zero and the prop was windmilling. *I subconsciously monitor the sound of the engine during flight and for once it was notably absent.

The airspeed was rapidly decaying so I immediately deactivated the autopilot and pushed the nose down

while simultaneously declaring an emergency.

Brussels control acknowledged the mayday and asked what I needed, I requested vectors to the nearest airport ‘Ostend is 10 miles north of your position’ so I started a descending left turn to the north.

Boost pump selected and left on, I switched fuel tanks.

Brussels control handed me over to Ostend.

I kept the airspeed >100kts as we had a lot of height to lose and didn’t need to stretch the glide.

Blue (prop) and red (mixture) knobs pushed fully in, the magnetos were left selected to ‘both’ and not touched throughout the emergency.

Speedbrakes deployed as still very high.

We crossed the north Belgian coast approximately 8000ft, I was offered but unable to accept an ILS for rwy 08 as engine not producing power - I requested and was approved for a glide approach. Soon after this we became visual with the airport but were still very high so I selected the gear down and flaps to T/O. I created a track on the G1000 extending the centreline of runway 08.

We were very high on initial final approach ?3000ft…

Passing 1000ft I selected full flaps and instantly (?possibly related to attitude change going from 10-33 degrees of flaps) the engine burst back into life. I put the landing lights on, completed final approach checks and made a surprisingly smooth landing on runway 08. We were met and escorted to the GA apron by the fire crew.

The engine sounded normal again until shutdown.

On external inspection… there was no water in the fuel.

But there was around 5mm of ice on the underside of the wings.

There was no visible ice on the leading edges or upper surfaces of wings so I had not been using the TKS deice system.

Also the fuel tank vents were unobstructed.

The oil quantity was ok.

Fuel quantity was also fine - there was approx 30usg remaining per side confirmed visually. We took the train back to Exeter while the aircraft was inspected by a local shop. The engine data from the G1000 revealed FF dropping to zero as the event preceding the engine failure.

The aircraft was inspected 11 days later (such is life in Europe!) - they found some evidence of ‘a few drops of’ water in the gascolator but nothing else amiss. Even though the aircraft was always kept in the hangar I had fully fuelled the aircraft in light rain the evening before the flight and had strained both wing fuel tanks prior to takeoff but I have to admit I omitted to pull the gascolator drain. As you can imagine I researched the subject of freezing avgas in some detail and found several reports of similar incidents affecting piston aircraft that were cruising in the flight levels in very cold conditions.

The Mooney POH refers to this risk and recommends adding anhydrous isopropyl alcohol 99% purity to avgas to make around a 1% solution in order to remove any free water, lower the freezing point of avgas from -57C and prevent these situations happening.

Another friend who suffered a similar power loss in his TB21 feels it’s more likely due to the fuel servo icing up within the engine and considers that this is less likely to happen when the TKS system is active as some of the prop TKS fluid gets into the engine bay and prevents the tiny nozzles from icing over.

Two fellow European Mooniacs generously gave me a lift from EGTE to EBOS to collect my aircraft yesterday and after adding isopropyl alcohol to the fuel and a thorough ground run-up I flew home to Exeter with no issue.

Freezing avgas/water was not on my radar but is a very real issue… I’ll be straining tanks and gascolator and adding IPA for long/high/cold sectors throughout the year and getting into warmer air as soon as possible if I ever encounter similar symptoms. Having seen the normally unseen underwing ice after landing I’ll also have a lower threshold for switching on the TKS deicing system in very cold conditions.

Anyway - I learned that freezing flight levels can bite…

 

 

Sent from my iPhone using Tapatalk

[Fly Boomer]

16 minutes ago, James McDiarmid said:

Engine Out 4/5/24

Glad it worked out.  Almost impossible to determine the cause since you cannot reproduce the conditions.  Suggest to always sump the wing tanks first, and then the gascolator.  If you do not have a means to collect and examine what comes out of the gascolator, drain each side for several seconds, push the airplane back a few feet, and examine the liquid on the ramp.

[M20TN_Driver]

That is a crazy scenario for sure. If you were indicating that the freezing point is -57c for a gas but you were flying at -32c why would you think the gas was frozen?

[EricJ]

1 hour ago, James McDiarmid said:

*I subconsciously monitor the sound of the engine during flight and for once it was notably absent.

This is a technique that I have also made use of. 

Also, kudos for handling it well and getting it on the ground safely.   

Where is the engine air inlet on the Acclaim-S?  If there was ice under the wings, did you note or recall whether there was ice on the engine intake?

The Continental engines don't have a fuel servo like the Lycomings, and use a different type of fuel injection system.   Either is susceptible to water freezing in a fuel supply line, though, which sounds like may be what happened.   Isopropyl alcohol is the usual avgas additive to help prevent this if one suspects water is present that can't be drained out.   

 

 

[N201MKTurbo]

Did you switch tanks?

Did the engine start working when you got down to warmer air?

[N201MKTurbo]

Only once in 6000 hours have I had a fuel injector ice up. It did what you said. The mixture kept getting lean until the knob was all the way in. Then the engine quit. I was IFR at 11000. I declared an emergency and requested an approach to Kingman Az. I was right over it. When I got down to about 6000 it started working again. I landed and spent the night. I was a bit shaken. 
 

The next day the sky was clear and the plane worked normally.

[James McDiarmid]

That is a crazy scenario for sure. If you were indicating that the freezing point is -57c for a gas but you were flying at -32c why would you think the gas was frozen?

Can’t be sure but… maybe it was a water/avgas mix, maybe the headwind/ice crystals brought it down further, it took 2h to happen so I’m sure temperature played a role


Sent from my iPhone using Tapatalk

[James McDiarmid]

This is a technique that I have also made use of. 

Also, kudos for handling it well and getting it on the ground safely.   
Where is the engine air inlet on the Acclaim-S?  If there was ice under the wings, did you note or recall whether there was ice on the engine intake?
The Continental engines don't have a fuel servo like the Lycomings, and use a different type of fuel injection system.   Either is susceptible to water freezing in a fuel supply line, though, which sounds like may be what happened.   Isopropyl alcohol is the usual avgas additive to help prevent this if one suspects water is present that can't be drained out.   
 
 

I didn’t see any ice around the cowl - just under the wings


Sent from my iPhone using Tapatalk

[James McDiarmid]

Did you switch tanks?
Did the engine start working when you got down to warmer air?

Yes to both


Sent from my iPhone using Tapatalk

[Paul Thomas]

1 hour ago, James McDiarmid said:

I didn’t see any ice around the cowl - just under the wings


Sent from my iPhone using Tapatalk

Is that common for the Mooney to only have formation under the wing? If so, that's scary to have it first adhere where the pilot cannot see it.

I'm glad you were able to get the airplane down safely. There is a lot of learning lessons there; it's not because one is not able to get the engine re-started that one should quit trying. I had not considered 100LL freezing either. 

Edited May 27, 2024 by Paul Thomas

[James McDiarmid]

Is that common for the Mooney to only have formation under the wing? If so, that's scary to have it first adhere where the pilot cannot see it.
I'm glad you were able to get the airplane down safely. There is a lot of learning lessons there; it's not because one is not able to get the engine re-started that one should quit trying. I had not considered 100LL freezing either. 

I don’t know… my guess is that the underside of the wing stayed iced up due to the rapid descent (10 minutes from FL230 to landing) - here’s a picture - both wings had identical icing


Sent from my iPhone using Tapatalk

[LANCECASPER]

14 hours ago, James McDiarmid said:

In retrospect the early indication of an issue was very occasional rough running which started 20 minutes or so before the engine stopped.

I appreciate you sharing all of what you did and the lessons learned. Great job getting it back on the ground safely.

The other lesson I picked up from your post is that if you hear or feel something that doesn’t seem right, like the occasional rough running engine, push NRST and Direct-To on the GPS and figure it out on the ground before the engine stops.

I learned to fly in very cold weather (North Dakota USA). At those outside temperatures that you were flying any roughness would make me first think water in the fuel. Your fuel didn’t freeze but the water in it did.   Our airplanes speak to us, sometimes listening sooner to what they are saying buys us the extra time we need. 

[Fly Boomer]

3 hours ago, Paul Thomas said:

I had not considered 100LL freezing either. 

This comes from a document prepared by the National Academies of Sciences, Engineering, and Medicine on airport refueling operations.  As you can see it says 100LL freezes at -72 F.  I doubt if it was that cold.  Probably water in the fuel system.

 

[N201MKTurbo]

I don’t think the ice on the bottom of the wing is typical airframe ice. I would call it frost from cold soaked fuel. It could have formed on the ground after you landed.

No matter how it got there, it doesn’t look thick or rough enough to affect the flying quality of the plane.

https://skybrary.aero/articles/hoar-frost#:~:text=Hoar frost may form on,fuel in the wing tanks.

[exM20K]

1 hour ago, N201MKTurbo said:

I don’t think the ice on the bottom of the wing is typical airframe ice. I would call it frost from cold soaked fuel. It could have formed on the ground after you landed.

No matter how it got there, it doesn’t look thick or rough enough to affect the flying quality of the plane.

https://skybrary.aero/articles/hoar-frost#:~:text=Hoar frost may form on,fuel in the wing tanks.

Right.  It is basically frost, and it will form only on the ground, requiring cold fuel from aloft and humid air on the surface.

-dan

[Brandt]

The POH references keeping the fuel pump on at low boost above FL180.  COuld this have been at least part of the issue?

[James McDiarmid]

The POH references keeping the fuel pump on at low boost above FL180.  COuld this have been at least part of the issue?

Yes… possibly… I have to admit that I hadn’t seen or read that recommendation so I didn’t have low boost pump on… having said that I have flown many hours at similar levels and not had a problem… it may have been a contributing factor


Sent from my iPhone using Tapatalk

[neilpilot]

Several of the replies above imply that the action of carefully sumping your tanks would have prevented the issue.  While it's obviously good practice to remove free water, there is ALWAYS some soluble water in Avgas.  Water solubility changes with fuel temperature, and the warmer the fuel the higher the dissolved water content.  Dissolved water will not appear in the sump, and is distributed throughout the fuel tank.

As you climb into low temperature conditions, the fuel in your tanks cools.  If sufficiently cold, the dissolved water comes out of solution and forms very fine ice crystal suspension that can collect in, and eventually bind, the fuel screen/filter.  If you're lucky, these crystals will melt as you descend into higher ambient temperature, as I suspect was the case for JM.

IPA, even in the form of Red Label HEET, is the most effective preventative measure for our Mooneys if you plan to fly into relative extreme cold conditions.

 

 

[exM20K]

@neilpilot I think this is likely what happened.  The Acclaim POH has this note:

It matters where you get your fuel.  I buy maybe 1/3 of mine on the gulf coast of Florida.  It probably has a lot of H2O in solution as compared to what I buy in Illinois.

make sure to get the good stuff (90%) isopropyl solution.

-dan

[GeeBee]

The ice on the bottom of the wing is environmental icing. Icing from a cold soaked fuel into moist air. That is why you see MD-80s de-icing in Florida in the middle of the summer, until wing heaters were installed.

With regard to being high on the approach, practice sometime with a safety pilot, gear down, flaps up, at best glide speed a full cross control slip. On a long body you can get about 2500-2700' FPM out of it. You can lose a lot of altitude really, really fast.

I too had a problem with soluble water in the fuel in the winter. 99% pure IPA works wonders, but it is hard on the tank sealant. Make sure when you add it you are adding it in the fuel stream as you refuel. Do not just dump it in the tank because you will find your tank sealant wearing fast below the filler port. I have found 1 to 1.5% is more than enough. (Thank you Paul Kortopates)

[dkkim73]

Are there any gotchas with sourcing the isopropyl alcohol? I haven't looked too closely at the drugstore stuff, not sure if there are small additives that might cause any issues. I figure it doesn't have to be anhydrous, just mostly alcohol.  The POH above warns us off some other auto gas tank additives. 

Seems like ordering fluids from aviation places dings you with hazmat fees, etc. 

D

[dkkim73]

On 5/28/2024 at 11:53 AM, Brandt said:

The POH references keeping the fuel pump on at low boost above FL180.  COuld this have been at least part of the issue?

The explanation for this (running low boost >FL180) I recall was to prevent vapor lock. And higher pressure might itself in theory help with small ice crystals. 

But, it also occurs to me that: 

-If this helps recirculate fuel through the system (e.g. running [high] boost pump with mixture cut off *I believe* recirculates fuel on a part of the system very near the throttle body but does not inject it, instead routing it back to the tanks... hence used to facilitate hot starts). Wouldn't this potentially help keep ice from forming in at least a part of the system? 

Curious if this forms any part of the rationale for the procedure. 

David

[kortopates]

It has to be anhydrous. Anhydrous simply means no water in. Rubbing alcohol is 70% isopropyl and 30% water. You're looking for 99% Isopropyl - there is no such thing as 100%. 90% isn't anhydrous either with 10% water. You can buy the 99% in the gallon from industrial supply outlets or you can buy it from many drug store outlets in  pint quantities.

[N201MKTurbo]

While you can dehydrate IPA to 100%. As soon as you open your apparatus to the atmosphere it will start absorbing water. So they just call it 99%.

It would be possible to dehydrate gasoline in the tank using a pervaporation membrane. You would need a vacuum source. You could mount a small vacuum pump in the plane, or just make a device you would stick into the tank while parked.

[EricJ]

59 minutes ago, dkkim73 said:

Are there any gotchas with sourcing the isopropyl alcohol? I haven't looked too closely at the drugstore stuff, not sure if there are small additives that might cause any issues. I figure it doesn't have to be anhydrous, just mostly alcohol.  The POH above warns us off some other auto gas tank additives. 

Seems like ordering fluids from aviation places dings you with hazmat fees, etc. 

D

I used to buy the "99.9% Ultra-Pure" (supposedly) stuff at Fry's Electronics, but we can't do that any more.   

You can get it on Amazon, though, if you can't find any locally.   Looks like many selections available.