Ben Visser on unleaded avgas

[PaulMillner]

There's an article by Ben Visser in General Aviation News about the unleaded fuel problem today! Interesting article. Care to comment, Paul?

Unleaded fuels and exhaust valve recession — General Aviation News

What needs to happen to move forward towards an unleaded future for general aviation?

generalaviationnews.com

A few years ago, I was giving a talk during a state aviation maintenance symposium when someone in the audience asked if there {were} any potential problems associated with the new proposed unleaded 100 octane avgas.
I explained that there could be a problem with an adequate rich knock rating for the fuel, which could lead to knock complaints in large engines and possible exhaust valve recession on some certified aircraft engines.

Not sure where Ben is going here... 100LL has a lean rating of 100 and a rich rating of 130. If Ben is saying that unleaded fuel *also* needs a rich rating of 130, he is correct. If, as some have said, he's saying unleaded fuel needs a HIGHER rich rating than leaded fuel to avoid detonation damage, that is erroneous... it comes from a math error (!) in the FAA's "ASTM" octane methodology. It's a sad thing, but goes to show that merely being an ASTM spec is not guarantee of merit.

At that point someone else raised his hand and stated that he read on the internet that exhaust valve recession was a complete myth and there was absolutely no risk of this occurring.
I then pointed out that in the 1980s an oil company in California looked over the ASTM D-910 specification for 80/87 avgas and noted that the spec only listed a maximum lead level of 0.5 grams per gallon and no minimum level. It then started selling an unleaded 80/87 avgas. There were no problems for a while, but then there was a rash of engine failures because of exhaust valve recession.
Many of these problems occurred in engines such as the Continental O-200 engine powering Cessna 150s used in flight training operations.
A well-known engine rebuilder from the Twin Cities then got up and shared several slides of engines that he had worked on that failed after using unleaded auto gas and had experienced exhaust valve recession.
Following this information, the person in the audience who said exhaust valve recession was a myth said maybe the engine rebuilder was right, but he could not believe that because the article he read was on the internet, so it had to be true.

I guess Ben is making the point that some folks believe everything they read. But, let's move on. More salient is that "lack of lead" problems in both mogas and avgas generally has come from not understanding the shift in delivered octane that occurs. More on that below.

Since that time, I have written several articles about the possibilities of exhaust valve recession once all of the lead is deleted from aviation fuel. These articles had a fair number of comments, including some from people who did not believe that it could happen. I also talked to engine manufacturers who felt that it was not a concern.
And then the University of North Dakota, which has a large well-maintained fleet of aircraft, started running on unleaded Swift UL94 fuel and guess what? They started seeing exhaust valve recession on several aircraft.

This is an exact parallel to the exhaust valve seat recession we saw in automotive engines in the 1980's that was associated with the phasedown of lead in mogas. But it wasn't the lead, it was the associated impact on fuel octane rating. Let me explain.
Blending gasoline in the 1970's, adding lead was the easiest and least expensive way to raise octane rating. But the octane engines used to measure octane rating were manually operated back then, and somewhat subjective. So two runs on the same sample could give an octane rating as much as 1 octane number apart. Also, the fine for selling gasoline that tested below the advertised number on the pump was as much as $1,000/gallon. So blenders would add extra lead, just in case. As a result, though regular unleaded might be labeled 87 octane on the pump, the actual octane of the fuel could be as high as 89 or 90 on any given day.
As the lead phasedown began in the 1980's it became more and more difficult to achieve octane... very expensive (platinum and rhenium catalyst) and energy intensive technology was required to boost gasoline octane. So blenders sharpened their pencils, and *delivered* octane began to decline while the number on the pump stayed the same.
Finally, the EPA accepted an industry proposal to exonerate blenders from government grab-sampled octane numbers that were too low, as long as the blender could demonstrate that they were in statistical process control (Edward Deming) of their blending process. This lowered delivered octane even more, to the pump stated number plus or minus a few tenths of an octane number.
As this process progressed, some engines without computerized controls began to show distress. The octane the engines were actually running on was declining, even though the number on the pump stayed the same. Fairly extensive work demonstrated this was so by the 1990's, but by then, no one cared anymore, as lead had been outlawed.
It appears the same thing happened to the UND engines... Although they were certified on 91/96 octane avgas in 1962, the Lycoming O320/O360 series engines hardly ever operated on that fuel. Most major blenders ceased making 91/96 by 1964. And, back then, no one leaned very much... they ran their engines rich. So the "91/'96" engines built most of their operating experience running on 100/130 or 100LL (that is rated 100/130).
At UND, however, all of a sudden these "91/96" engines were operating on 94 octane instead of 100 octane. And UND's practice is to lean to peak in cruise, which offers much less detonation resistance than the rich mixture settings of old. It's not surprising that engines began to suffer distress when the delivered octane declined (by six numbers) and the operating conditions were more severe than during the certification of those engines sixty years ago.

So why is this happening at flight schools and not in private aircraft operated on Swift fuel or unleaded auto gas?
The big difference is the single source of fuel.
Many private aircraft were broken in on 100LL, plus many are used on cross-country flights, where they are refueled with 100LL occasionally. Most flight schools do all of their refueling at only one location, so most aircraft in a school's fleet never see any lead in normal operation.

That's kind of clumsily stated. But again, it's not the lead, it's the octane rating of the fuel that is material.

The reaction to the UND valve recession problem was interesting. Lycoming thought it could be due to the aromatic content of the Swift fuel.
But guess what? Most 100LL blends have contained aromatics since the early 1970s.
The only difference between 100LL and the Swift fuel is - wait for it - lead.

The Swift 94UL samples I've seen are only about 2% aromatics. 100LL is often 10% to 20% aromatics. And... there's no research that associates aromatics with valve recession. Lycoming's rationale makes no sense. Besides, much 100/130 avgas also contained aromatics, and has since WW2.

{continued}

 

[PaulMillner]

Part 2:

What are the main factors that prevent exhaust valve recession?
In the 1970s auto manufacturers found that hardened exhaust valve seats in liquid-cooled engines equipped with knock sensors helped prevent the problem.
The problem was also associated with load and the RPM that the engines experienced.

Either technology, hardening the valve seats *or* retarding the timing to reduce intracylinder pressure, would resolve the problem.

The problem for general aviation is that aircraft operate at high load, high RPM, with air cooling, and high exhaust valve and seat temperature - all factors that are in the danger zone.

What it comes down to is that the way the engines are operated, the engine octane requirement is greater than what was supposed during the 1962 certification effort. It's not surprising that our understanding, and instrumentation, has improved in sixty years.

Some of the discussion about the problem at UND included the thought that the university's planes were operated on the lean side of stoichiometric, which would aggravate the problem.
But I believe they have multi-point temperature probes on all cylinders and they lean based on the monitor.

Actually, running lean of peak (stoichiometric) would reduce the problem. But UND claims that they run at peak, which is cooler than running 30F to 50F rich of peak, but hotter than running at 100F+ rich of peak like the old days.

There is also a very real possibility for knocking on these engines using the 94 octane Swift fuel. Almost any knocking will raise the temperature of the exhaust valve and seat significantly and can lead to exhaust valve recession.

That is all true. In fact, even before you get to detectable knock, intracylinder pressure during the combustion cycle is rising, and this presses the exhaust valves harder and harder into their seats, encouraging micro-welding of the valve to the seat to occur. When the valve subsequently opens, the tiny welds break, making the seat and valve rougher, reducing heat transfer that is supposed to cool the valves, and setting the state for more micro-welding on the next combustion cycle.

A multi-point EGT or CHT system may be mandatory for all aircraft before they are operated on an unleaded fuel.

Engine monitors are a very good thing; I'm not sure whether there's lead or not in the fuel is germane. If you're going to reduce the delivered octane of the fuel, though, then it does become more important to monitor what's going on.

In addition, knock sensors would help, but are not too practical in the aviation world because of individual cylinder assemblies.

Individual cylinder assemblies really don't have much to do with it. Knock sensors can be added to individual cylinder assemblies, both Lycoming and Continental have demonstrated that. The historic challenge was differentiating the knock noise signature from all the ambient spectrum-wide noise from propeller blast. But setting an aviation engine up to retard timing to avoid knock is not an entire answer... first, you have to prove that even if you retard the timing you can still deliver the certified horsepower, or the aircraft becomes unairworthy. This is a big deal.

What else needs to be done before we switch over to all unleaded fuel?
There are possible engine modifications, such as liquid cooling, but that would be a very high cost for every general aviation airplane out there.
I think the best answer would be to put an additive in all unleaded fuels to reduce or eliminate the problem. Something like MMT - Methylcyclopentadienyl manganese tricarbonyl, a fuel octane enhancer produced by Afton Chemical Corporation - might work.
The problem with MMT is that at normal treat levels the additive produces dark red whisker deposits on the spark plugs and other surfaces. These deposits can cause fouling and other problems.
But at very low levels it may not cause any problems and may prevent the recession problem.

Yeah, MMT is a non-starter. Both Phillips 66 and Lyondell/VP-Racing thought MMT might be part of the solution. Phillips' fuel failed in testing, and they've "paused" their development effort. (Remember Shell's unleaded fuel? It didn't use MMT but a different problematic constituent. It's been paused since January 2021. Shell said they would be back after the EPA endangerment finding. That happened six months ago. So far, no Shell...)

What needs to happen to move forward towards an unleaded future for general aviation?
A repeatable valve recession test to determine if MMT or other additives - such as TCP, an FAA-approved aftermarket product that helps prevents lead fouling on valves and spark plugs - will eliminate possible exhaust valve recession.
But like lead and other fuel additives, TCP has its own health problems, including irritation of the eyes, skin and respiratory tract, drowsiness or dizziness, and depression of the central nervous system.

I'm not aware of any pathway by which TCP (or it's replacement TPP) would prevent valve recession. Lycoming advocates TCP/TPP in *oil* to provide cam lobe and tappet face scuffing protection. But valve seat wear isn't a scuffing phenomenon, it's an overpressure phenomenon from inadequate fuel octane. And by the time any TCP/TPP in the fuel got to the valve seat, it would already have been combusted, unlike the oil application. TCP has been added to fuel to better scavenge lead, but that's not applicable to unleaded fuels.
On the contrary, what needs to happen has already happened: GAMI has come up with a fuel formulation that offers adequate octane rating to prevent problems. In fact, GAMI's fuel rich rating (the 130 number of 100/130) is over 160, providing huge headroom to prevent detonation related difficulties. Unfortunately, both Lyondell and Swift Fuels are advancing ether-based fuels instead. And both have said that their fuels will NOT meet the needs of all the engines in the fleet. That seems like a fatal flaw. GAMI's fuel has no such limitation.

Paul

 

[Capt. Geoffrey Thorpe]

"And then the University of North Dakota, which has a large well-maintained fleet of aircraft, started running on unleaded Swift UL94 fuel and guess what? They started seeing exhaust valve recession on several aircraft."

My understanding is that this is not necessarily true. One valve on one engine on a twin had definite valve recession. Other engines may have suffered valve recession. The cause is not clear. Lack of lead? Octane? Measurement issues? Or, they had just never looked before...

(based on information from Mike "one word at a time" Bush)

 

[NordicDave]

(based on information from Mike "one word at a time" Bush)

Hilarious! I thought I may have been the only guy to listen to his EAA webinars at 1.5-2x speed!

 

[Jim K]

Hilarious! I thought I may have been the only guy to listen to his EAA webinars at 1.5-2x speed!

Nope....

 

[Clip4]

“The problem for general aviation is that aircraft operate at high load, high RPM, with air cooling, and high exhaust valve and seat temperature — all factors that are in the danger zone.”

2300- 2700 RPM is high?

 

[tspear]

“The problem for general aviation is that aircraft operate at high load, high RPM, with air cooling, and high exhaust valve and seat temperature — all factors that are in the danger zone.”

2300- 2700 RPM is high?

For a ship diesel engine, yes it is.

Tim