Aviation Alternative Fuel Compatibility Issues

This is a very complex question.  There are three big parts to it:  (1) a huge legacy fleet of aircraft with components known to react badly to alcohols,  (2) pilots not trained to use the different fuels which really do respond quite differently in so many ways,  and (3) an already-disruptive transition from low-aromatic higher-leaded 100/130 grade aviation gasoline to the high aromatic lower-leaded 100LL grade of aviation gasoline.  During that transition to 100LL,  many of the same types of seal failures were seen as are seen converting to ethanol-containing fuels. 

With ethanol already in motor gasoline,  usually at the 8-10% level,  exposures have already occurred with reported impacts that vary widely.  This is for the airplanes that can use the motor gasoline STC.  That STC actually prohibits the use of ethanol blends,  which are now about all that is available for motor gasoline. 

Therefore,  that conversion is not so popular now,  in part because of the prohibition,  and in part because the problems encountered when “using the ethanol-containing motor gasoline anyway”.  These problems trace to materials in common aviation use that are incompatible with ethanol.  I have grouped them in the discussions below.  For non-aviation items,  see "An Update on Ethanol Fuel Use" dated 11-2-13.  

See Update 9-26-16 at end of article.

Metals

Intergranular corrosion of aluminum alloys is worse with the aluminum-copper Duralumins.  These are the primary structural materials in aircraft,  including aluminum tubing for fuel lines.  That is why “wet-wing” tanks and aluminum tube fuel lines need a surface coating to resist alcohol-induced corrosion.  Unfortunately,  the Alodine process usually recommended for these items requires their removal from the aircraft,  a labor-intensive and expensive thing.  This could be remedied by changes in production for new aircraft,  but the legacy fleet is very large,  and it has changed little with the passage of decades. 

You would be better-off using a compatible bladder-lined tank,  and Alodining just the aluminum lines (which raises fuel bladder compatibility as discussed below). 

Some of the steels exhibit some corrosion sensitivities,  too,  but not nearly as bad as the Duralumins.  The zinc casting alloys seem more sensitive to steam than alcohol,  as regards intergranular corrosion. 

Non-Duralumin aluminum casting materials for carburetors and similar devices are less sensitive to these effects.  These generally hold up rather well with ethanol.  I have several from air-cooled engine cars that are just fine after years of calendar time,  and hundreds of hours of operation,  with E-85 ethanol. 

Metering Devices and Related Items

The real issue about fuel metering devices is twofold:  (1) seals and (2) small passages.  Seal compatibility is discussed below.  There are potentially-serious problems with small passages,  whether the fuel is gasoline,  alcohol,  or blends of them.  When gasoline (motor or aviation) evaporates,  it leaves behind a coating of gum and varnish on the solid surfaces (motor gasoline is far worse,  but both do it).  These deposits can plug small passages,  as anybody who has ever cleaned-out a “dirty carburetor” on a car or lawnmower already knows.  It can happen to aircraft metering devices,  too.

The problem is slightly different when ethanol evaporates inside a metering device.  It leaves no gum and varnish,  and the corrosive effects of direct liquid contact are not the bigger source of troubles.  It is the ethanol vapor that attacks the aluminum,  leaving behind a fine gritty aluminum oxide,  and a low-density,  “sticky-gooey” aluminum hydroxide gel-like material.  The grit is a really big wear problem for any moving parts,  and the gooey gel plugs up the passage. 

Oddly enough,  gasoline turns out to be a decent solvent for the ethanol-induced deposits,  and ethanol is a decent solvent for the gum and varnish that gasoline leaves behind!  That raises the hope that blends of the two might behave better than either fuel “neat”,  but no one really knows. 

The real point is that engines should not be parked without use for months at a time,  with fuel in the fuel delivery system.  Bad things will happen with either fuel.  If you know the engine will not be run for a long time,  the better choice is cut off the fuel to it at the tank valve,  and drain the lines,  pumps,  and metering device dry.  That’s true with any fuel,  and therefore very good advice.

Composite Structural Materials

The most common choices are epoxy- and polyester- or vinyl ester-based materials.  The choice of carbon or glass fiber really isn’t the issue,  it is degradation of the matrix polymer that can lead to softening and failure.  Many epoxies are good with gasoline wetting,  so that one may build a “wet wing” fuel tank with it.  Some of these are OK with ethanol,  others are not. 

You have to “dunk test” a sample of your specific composite material to find out.  The softening of the epoxy will show up in a couple of days to a week.  The same is true of polyesters and vinyl esters.  You simply must check before risking ethanol exposure (or gasoline for you homebuilders).  If you must use ethanol,  and you find your matrix is susceptible,  then you must use a fuel bladder,  and it had better not ever leak!

Sealing Materials and Non-Metallic Components

Neoprene,  ethylene propylene,  and Buna-S work well with ethanol,  just as they do with gasoline.  The fluorosilicones and the Vitons are not compatible with ethanol.  Nobody knows for sure about the Buna-N materials.  These are all materials used in O-rings and similar seals,  throughout the aviation fleet.  The polysulfides often used as fuel tank sealants don’t work with ethanol.  (Some of these same elastomer seal material problems cropped up with the transition from 100/130 to 100LL,  due to the high aromatics content,  which is also a corrosive solvent.)

There are some nylons that react badly to ethanol by swelling and weakening.  Teflon is pretty much impervious to everything.  The Nitrophyl and cork float materials are incompatible with ethanol,  although polypropylene and polyethylene seem suitable.  All sorts of fuel metering,  filtering,  and pumping is done with devices incorporating these materials in their working parts. 

Polyurethane (foam or otherwise) is not compatible with ethanol.  No one knows about the Acetal polymers.  These can be found in some fuel tank gaging devices. 

Fuel Bladders

Neoprene works,  polyurethane does not.  Polyurethane is by far the most common fuel tank bladder out there in the aviation fleet.  Very unfortunate for those wishing to do an ethanol conversion!

Polar Solvent Effects in Pumps

Once the incompatible shaft seal material fails in an electric fuel pump,  the fuel can leak into the electric motor itself.  Nonpolar gasoline does not cause electrical arcing problems,  polar ethanol does.  Both are considerable fire hazards when exposed to electric sparks.  You have to avoid this with the right seal material in the first place.  The problem is the legacy fleet and the store of parts that supports it:  many or most of these devices are not compatible with ethanol fuel. 

Capacitative Fuel Gaging Devices

These cannot work when the fuel is a polar solvent.  Ethanol is polar,  gasoline is nonpolar.  Ethanol fuel doesn’t work at all with capacitative gaging,  and neither do ethanol-containing blends. 

Changes to Operating Characteristics and Pilot Training

These depend upon how the conversion was done,  and the nature of the fuel or blend to be flown.  I know about the characteristics of some of the conversions,  but not all of them.

                Cold start

“Neat” denatured ethanol meeting ASTM specification D-4806 “Fuel Grade Ethanol” is 95% ethanol,  5% gasoline or petroleum alkylate,  and under 0.5% water.  It will not cold start below about 50-60 F ambient temperature,  because the volatility is too low (very low vapor pressure,  a high latent heat of evaporation,  and essentially a constant boiling temperature instead of distillation curve behavior).  To use it requires re-plumbing the engine start primer to a separate canister containing some gasoline. 

This separate start canister change can be avoided if the fuel contains 15-or-more% gasoline instead of 5%,  as in automotive E-85 and aviation AGE-85.  The automotive grade isn’t really 85% ethanol in winter.  It can be as low as 70% ethanol,  and is usually about 75% ethanol, during winter,  for even better starting than “real E-85”.  AGE-85 is specified to be 85% ethanol,  14% gasoline,  and 1% biodiesel. 

                Mixture Control

The conversion of just about any aviation carburetor,  and the Bendix RSA-5 series of injection devices,  leads inherently to a flex-fuel system capable of metering all the intermediate blends.  I honestly don’t know about conversions of the other fuel injection systems. 

For a carburetor,  you want sufficient flow capacity for ethanol in both the main jet and idle systems,  at otherwise the same air flow pressure signals.  Basically,  the jets get bigger.  Because it is an aviation carburetor,  the pilot has (and is expected to use effectively) a mixture control.  It is a change in the operation of this mixture control that the pilot will see;  the rest appears unchanged to him. 

The same thing is true for Bendix Airmotive RSA-5-based fuel injection.  You need to drill out the idle valves slightly for greater idle flow capacity,  and you use enlarged fuel injectors at the cylinders,  all at the same regulated fuel line pressures as before. 

When operating on ethanol with a properly-executed conversion,  “full rich” will be all the way forward with the mixture control,  just like it was before.  If you pull it all the way lean,  the engine will die for lack of fuel,  just like it did before. 

On gasoline,  however,  “full rich” is only about halfway forward.  If you go all the way forward,  the engine will die from overrich mixture.  All the way back is still lean-out,  just as before.  Only the sensitivity is different:  a shorter travel of the control gets the full effect.  The risk is the lack of a mechanical stop for “full rich”.  

On intermediate blends,  “full rich” will be between those forward and halfway positions,  more or less linearly proportional to the blend.  Mishandling the control can cause the engine to die from overrich mixtures.  Greater pilot awareness of,  and a gut feel for,  how the engine is running is simply required.  This does take experience with engines;  it is not for the novice.  But,  it is not hard,  and can be learned quickly.

Ethanol burns differently than gasoline (soot-free flame),  and can be leaned far more aggressively than gasoline.  Blends fall somewhere in between the “neat fuel” extremes. 

With gasoline,  you need to enrich the mixture very slightly from the “perfect” mixture point in order to control hardware temperatures and avoid engine damage risks.  This usually takes the form of finding the peak exhaust gas temperature,  and then enriching slightly to drop that exhaust gas temperature by some small margin,  usually around 50 to 100 degrees F.  You can just “peak” the sound of the engine,  and then bump the mixture control a tad richer from there,  and achieve essentially the same result,  just not as repeatably. But it works just fine,  all the same.  

With ethanol,  hardware temperatures will be lower for the just about the same exhaust gas temperatures.  This is because of the reduced heat load on engine parts from the adjacent flame,  since there is little soot in the flame to radiate.  You can simply find the peak exhaust gas temperature,  or even just “peak” the sound of the engine.  Such mixtures are nearer the excess-oxygen point,  but the hardware is running cooler,  which acts to offset the oxygen risk. 

                Fuel Samples and Water Bottoms

With ethanol in the blend,  you will never see water bottoms,  which is completely at variance with all pilot experience operating on gasoline fuel.  With ethanol in the blend,  any such water bottom in the tank goes into solution,  right up to the phase separation point. 

If separation occurs,  the water and all the ethanol go into the bottom layer,  being denser.  The now-dry hydrocarbon floats on top.  Any dye in the fuel goes with the hydrocarbon.  It is easy to see the layers,  even if both components are clear.  But your typical tank bottom sample will be clear with phase separation in the tank,  when the 100LL (or 100LL blend with ethanol) ought to be blue. 

Very simply put,  do not ever fly with a phase-separated tank!  That boundary gets shaken up in flight,  so that you draw globules of one layer or the other,  not the mixture,   into the engine.  Power surges and possibly detonation can result. 

Unfortunately,  predicting that phase separation point is difficult at best.  But there is a simple go/no-go field test.  It uses the clear-tube fuel quantity gaging tube that most aircraft now have. 

                Is My Tank Separated As It Is?

If you gage a tank with the tube,  you essentially take a “core sample” of the fuel in the tank.  Experience shows that any separation boundary in the tank is preserved in the “core sample” taken with the gaging tube.  You can see it.  If you do,  drain it down to one layer (the dry 100LL gasoline),  and refuel on top of that.  (If you are using automotive gasoline,  drain it all and simply replace it with fresh fuel.  It has lost its ethanol content,  which was a large part of its octane rating.) 

                Will My Tank Separate Upon Refueling?

Your tank’s fuel blend may not be separated,  but still might separate if you refuel on top of it with one or the other fuels used neat.  There is a way to determine that risk,  which requires no knowledge of the blend proportions in your tank,  or in the fuel that you will add.  It does require than you can quantify the number of gallons still in your tank and how many gallons you are about to add.  In other words,  you need to know the “calibration” of your fuel quantity indicator. 

Use your bottom sample and a sample of what you propose to add,  in the closest approximation to the volume ratio that will obtain if you refuel.  Shake it up and watch it for separation for a couple of minutes.  If your sample separates,  so will your tank,  so don’t refuel that way.  Instead,  drain down and refuel with one fuel.  If it doesn’t separate,  neither will your refueled tank,  so top off your tank and fly on happy. 

                What If the Fuel Is “Old”? Does It Get Too Wet?

It is particularly important to investigate whether your tank is separated,  and whether it will separate upon refueling,  if the aircraft has sat idle for months with an ethanol blend in it.  

Aircraft tanks are vented,  and can accumulate moisture from the atmosphere in the air space on top of the fuel,  as the tank “breathes” every 24-hour cycle.  (With straight gasoline,  this is where water bottoms come from.)

The so-called azeotrope mixture of straight ethanol fuel is 5% water.  That is an unrealistic upper bound for what a real vented tank can absorb,  even over a very long period of time.  More realistic results are 1-2% water after 6 months to a year sitting,  even in the very humid air along the Texas Gulf Coast. 

I’ve never seen more than 2% water myself.  It’s usually no more than 1% around Waco McLennan County,  even after a year sitting.  If the aircraft sits idle that long,  the really significant risks are fuel evaporation deposits in the metering device as described above,  not water absorbed in the fuel. 

Putting straight 100LL on top of E-95 ethanol residuals with 2% water can indeed pose a refueling separation risk,  sometimes.  Sometimes not,  maybe quite often not.  The point is,  you simply do not know.  It depends upon how much fuel gets added compared to how much is still in the tank and how wet that has become.  That’s why the bottom sample prediction test described just above is so important when operating with blends. 

In the extensive experimental blend fuel work that I have done with cars,  blends near E-30 to E-40 typically require more than a 15-20% water addition to force a phase separation.  I typically use a 30-35% water addition when I run the forced separation test checking blend strengths,  just to make the test fully reliable.  It’s pretty much the same behavior at E-10 levels,  and at E-85 levels. 

                Power and Economy

Different investigators report different outcomes from power tests on ethanol versus gasoline.  It depends on whether you aggressively-lean with the ethanol,  as discussed above.  If you do,  the lower-compression engines may show around 5-10% better power and efficiency on ethanol.  The higher-compression engines may show 10-20% better power and efficiency on ethanol.  But you must aggressively lean,  or you won’t see much of this effect.  Maybe none. 

If you don’t aggressively lean and thus you don’t see the power and efficiency improvement,  then your fuel flow rates will essentially match the volumetric heating value ratio of the fuels.  Heating value is proportional to the ideal or stoichiometric air fuel ratio by mass,  so for gasoline (14.5:1) versus ethanol (9:1),  your ethanol flow rate will be about 14.5/9 = 1.61 times larger than your gasoline flow rate,  at otherwise the same power setting and conditions. 

If you do aggressively lean the ethanol (and you safely can),  then for a power improvement of 10% at lower compression,  you get an ethanol/gasoline flow ratio near 14.5/(9*1.10) = 1.46.  If you have a power increase of 20% at higher compression,  then the flow ratio is near 14.5/(9*1.2) = 1.34.  That sensitivity to leaning strategy is why different investigators using different leaning strategies get such widely-varying results. 

Blends are going to fall somewhere in-between,  and quite probably not in a linear fashion with blend ratio.  The leaning strategy also needs to vary with blend,  as there is more and more hardware-heating soot radiating in the flame,  as gasoline content increases.  To the best of my knowledge,  those tests to optimize leaning strategy with blend,  have not been done.  The risk is engine damage (usually burned valves and seats) if you get it wrong.  It takes a while to incur the damage. 

Fuel Density and Gross Weight

Gasoline (motor or aviation) is typically variable in density,  but usually about 6.1 to 6.2 lb/US gallon.  E-85 is pretty close to 6.5 lb/US gallon,  E-95 a little bit denser still.  That’s not much of a density difference,  and therefore not much of a risk for gross weight,  unless you are already (or habitually) very near the maximum limit.  Just use the tank volume and the higher blend density to figure the higher fuel weight.  If that puts you over gross weight limits,  then you offload a little cargo or fuel.  Simple as that. 

Concluding Comments

There’s enough risks and difficulties associated with using ethanol or ethanol blends in the aviation fleet,  that you probably don’t want to do this unless you have some other compelling need.  The most difficult risks are materials compatibility problems.  Corrosion can occur,  and must be mitigated.  Doing the conversion itself is not very hard,  nor is learning how to fly with it. 

Don’t do this at all if you have a capacitative fuel gaging system.  Period. 

One of the biggest issues is fuel flow capacity,  when ethanol requires substantially-higher flow rates.  Most of the time,  you will simply have to convert gravity flow systems to pump-fed systems.  Just up-sizing the fuel lines is nearly always inadequate by itself,  although it still needs to be done,  even with adding fuel pumps.  The convenient time to Alodine the fuel lines is when you up-size them,  though. 

You will have to select pumps and check valves with ethanol-compatible seals,  namely neoprene.  If you don’t pay close attention to this,  you will get into trouble later when the seals leak,  particularly in an electric fuel pump.  Check for neoprene diaphragms and polypropylene plastic parts inside engine-driven fuel pumps,  and inside carburetors or RSA-5 devices. 

Teflon-lined fuel hoses on the engine are the best choice,  although neoprene or Buna-S rubber is OK.  Be sure any fuel bladder is neoprene,  not polyurethane.  You also need to check the compatibility of the materials in your mechanical fuel quantity indicator (see “horror story” below).

Be sure your fuel sample device is polypropylene.  And be sure to acquire a tank-gaging sample tube,  especially if you intend to fly blends. 

The available STC’s are for either gasoline or ethanol,  not blends.  To the best of my knowledge,  blends are still only allowed in experimental category.  If you are experimental,  you have more latitude to use automotive fuel components,  which are generally a lot more ethanol-compatible these days.  Experimental guys therefore have it a little easier. 

                “Horror Story” about Fuel Quantity Indicators

There are two structural transparency plastics:  Lexan and Plexiglas.  Both are acrylic plastics,  just different manufacturers.  Long-term liquid ethanol contact can cause surface crazing in both of them (a web of tiny cracks).  But,  they are both quickly and completely destroyed by warm ethanol vapor contact.

When converting an old Piper “Pawnee” to use E-95 ethanol,  I ran across a mechanical fuel gage under a clear blister,  which was also the vapor collection space for the tank vent.    This works fine with gasoline,  but repeatedly failed in about half an hour with ethanol on a nice warm day.  The blister literally collapses into a wadded-up mess.  (It’s not a cheap item,  either.) 

That blister had an integral flange with drilled holes.  It bolted directly to the airframe on top of a paper gasket.  I had to replace it with glass,  which you simply cannot bolt down that way,  and for which custom blown parts are very,  very expensive. 

The solution was a steel base plate to which the ring lid of a Mason jar was welded.  A little silicone adhesive on the threads of the Mason jar sealed it to the ring lid.  This worked,  and is safe,  as demonstrated by test to the FAA. 

By putting the tailwheel in a chair to level the deck,  we made paint marks on the jar as we filled the tank.  That way,  we actually did end up with a more-accurate fuel quantity indicator than the original,  which was nothing but an inaccurately-positioned decal.

And that’s why there is a grocery store Mason jar in the STC for using ethanol in a Piper Pawnee!  

Update 11-4-13:

Taking into account my automotive,  small engine,  and aircraft experiences,  I would assess things as follows:

The automotive industry has long adopted materials compatible with high aromatic and ethanol content in gasoline.  That plus flex-fuel cars has driven the supporting parts industry in that direction for decades.  As a result,  you can reliably use blends or convert to E-85,  pretty much in anything from modern to very old.  And I do mean many decades old!

It appears to me that the 4-stroke small engine folks have also largely made the transition to ethanol-compatible materials.  I’m not so sure about the 2-stroke small engine folks,  and I have not run any stiff blend experiments in 2-stroke because of my preconceptions regarding lubrication.  E-10 seems to be OK in 2-stroke,  though,  and based on that,  I don’t have much concern about E-15 (it’s just not that different). 

I don’t think the boat motor folks have made the transition to ethanol-compatible materials yet,  not even on the 4-stroke side of the house.  It’s boat motors and 2-stroke equipment that I hear the most horror stories about.  (I haven’t really investigated this,  so my perception is only just that: a perception.)  If there was going to be a problem with water bottoms,  I would expect to see it in boats,  though. 

The light aircraft industry is still using the same materials they were using half a century (or more) ago.  Many of these are not even good for high-aromatic gasoline,  not to mention ethanol.  Commercial aircraft are today nearly all turbine,  for which biodiesel content in the jet fuel is proving to be a good fit (that’s a another whole topic area not discussed here).

Bottom lines:

Automotive:  feel free to experiment with stiff blends and neat ethanol.  The materials in common use for the last few decades seem to support ethanol compatibility,  by and large.  No mods necessary up to about E-42,  although I don’t recommend over E-35 because of minor cold start problems.  Conversions to straight E-85 are easy,  but will be successful only if you do all 3 required items (mixture,  timing,  extra intake heat).

4-stroke small engine:  feel free to experiment with stiff blends up to E-35 without mods.  I’d not recommend neat ethanol or E-85 conversions,  because the carburetors do not have removable jets and are therefore so very hard to modify.  It’s very hard to change the timing in a magneto ignition,  too.  The supporting parts seem to me to be largely ethanol-compatible in recent years.

2-stroke small engine:  I personally would not experiment with stiff blends in these,  because of lubrication fears (oil-in-fuel,  with an oil solvent also in the fuel?),  and because the supporting parts do not yet seem to be very ethanol-compatible.  Wait till the supporting parts industry has made the transition. 

4-stroke boat:  you can try stiff blends in unmodified engines,  but be aware that supporting parts may well be incompatible with ethanol.  You will have to determine the neoprene/polypropylene issues for yourself.  I’d rather wait until the supporting parts industry has made the transition,  before I went above about E-15.  If your parts are compatible,  up to E-35 should work just fine. 

2-stroke boat:  same as 2-stroke small engine,  and for the same reasons. 

Gasoline airplane:  you need an otherwise-compelling reason to do ethanol blends and conversions,  and you will run into incompatible materials and metal corrosion problems everywhere you look.  But it can be done.  And it does work. 

Turbine aircraft:  out of scope here.  Biodiesel blends do work just fine,  though.  

Update 9-26-16:

              From Biofuels Digest 9-25-16:

              DOE Study

In Colorado, a study conducted by DOE’s National Renewable Energy Laboratory (NREL), found that the petroleum components of ethanol-blended gasoline become degraded and unfit for use in an engine long before the ethanol portion takes up enough water to cause phase separation in the fuel tank. “Phase separation” occurs when an excessive amount of water is introduced into the fuel tank leading the ethanol and water to mix and sink to the bottom of the tank. In other words, gasoline becomes “stale” and unusable before water uptake by the ethanol component becomes a concern.

As part of the study, NREL scientists stored gasoline-ethanol blends ranging from E0 (0% ethanol) to E85 (83% ethanol) in actual lawn mower fuel tanks over several months in a climate-controlled chamber meant to replicate hot, humid environments like Houston and Orlando. The samples were tested at regular intervals for evidence of gasoline weathering and water uptake. In every case, the hydrocarbon components of the fuel became unfit for use in an engine before water uptake became a concern.

For gasoline-ethanol blends, it often took more than three months for phase separation to occur, meaning the fuel had already weathered to a point it was unusable. “In a small engine fuel tank in a constantly high-temperature, high-humidity environment, it takes three months or longer for E10 and other ethanol blends to take up enough water for phase separation,” the study found. “This confirms the statement by Mercury Marine that water uptake in E10 blends ‘…does not happen at a level or rate that is relevant.’”

President and CEO Bob Dinneen offered the following comments on the new study:

“Simply put, critics who continue to suggest E10 is a problem for small engines and boat motors are all wet. This research from NREL clearly demonstrates that gasoline goes bad long before the ethanol in the tank could cause any problems due to moisture uptake.

“Every manufacturer of small and off-road engines has approved the use of E10 in their equipment for many years. If owners of this equipment simply follow the manufacturers’ recommendations for fuel, maintenance, and winterization, they won’t have any issues at all. But, as this study shows, letting gasoline sit in your tank for extended periods of time is likely to cause some issues—irrespective of whether the gasoline contains ethanol or not.”

My take:  the water problems that small engine and boat motor owners report are due to the item being left out in the rain,  not the humidity.  The fuel tank caps have air vents.  The rain gets in that way.  Plus,  many folks are using fuel more than 3 months old.  Both a recipes for bad performance and maintenance problems,  even with zero ethanol in the fuel.