100LL: EPA proposed endangerment finding

[Fly Boomer]

48 minutes ago, jaylw314 said:

"Correlation does not mean causation" is only correct most of the time.  There has to be a point of sufficient evidence that the strength of correlation does mean causation for all intents and purposes.  That's how science works, otherwise, we'd still be banging the rocks tog  ether.

Not exactly.  Science is proving that A causes B, and that A is followed almost always by B, and that B rarely or never occurs without A.  "For all intents and purposes" is not proof.

[midlifeflyer]

6 hours ago, A64Pilot said:

If you look at aircraft engine history, pretty much all of the attempts to modernize them has failed, even when tested and built by actual engine manufacturers.

Where are all the geared high RPM liquid cooled engines now?

You mean like the Jet-A using, liquid cooled turbo, geared-prop in the newest DA-40S? Well, not particularly high RPM.

My only point is that the handwriting has been on the wall for a long time (we went from the Wright brothers to the moon in 60 years) but ignored. That's pretty much fact. Our interpretations might be different.

Edited October 17, 2022 by midlifeflyer

[jaylw314]

2 hours ago, Fly Boomer said:

Not exactly.  Science is proving that A causes B, and that A is followed almost always by B, and that B rarely or never occurs without A.  "For all intents and purposes" is not proof.

Edit: Sorry, that wasn't nice, I'll try to tone down the snarkiness

That is not how science works, sorry.

Edited October 17, 2022 by jaylw314

[1980Mooney]

7 hours ago, N201MKTurbo said:

I couldn't agree more. The regulation should be that 100LL cannot be discontinued until there is an equivalent supply of a substitute readily available.  

It seems like most here would use it if given the choice.  Market forces are more palatable than government mandates. 

Once it is available market forces will move fast.  100LL will disappear quickly.

I have said it before:

• AvGas sales nationwide are only about 420,000/gallons per day
• That sounds robust but there are about 3600 FBO's nationwide
  • That means the average FBO only sells about 116 gallons per day - that is basically two (2) 55 gallon drums
•  The average FBO cannot afford to invest in infrastructure to sell multiple grades of fuel - especially if one is unleaded and totally separate from the other.
  • Another $100,000+ investment for separate tank, truck pumps just to split the sales supporting maybe 50/ gallons per day?
  • If you amortize that over 5 years and add about $5,000/year for Maintenace and insurance that is about $25,000 spread over less than 20,000 gallons 100LL per year.  So the FBO adds another $1.25 or more to the cost of 100LL.
    • That makes the cost of 100LL about the same as G100UL - GAME OVER.

 

Edited October 17, 2022 by 1980Mooney

[ArtVandelay]

Once it is available market forces will move fast.  100LL will disappear quickly.
I have said it before:

• AvGas sales nationwide are only about 420,000/gallons per day
• That sounds robust but there are about 3600 FBO's nationwide
  • That means the average FBO only sells about 116 gallons per day - that is basically two (2) 55 gallon drums
•  The average FBO cannot afford to invest in infrastructure to sell multiple grades of fuel - especially if one is unleaded and totally separate from the other.
  • Another $100,000+ investment for separate tank, truck pumps just to split the sales supporting maybe 50/ gallons per day?
  • If you amortize that over 5 years and add about $5,000/year for Maintenace and insurance that is about $25,000 spread over less than 20,000 gallons 100LL per year.  So the FBO adds another $1.25 or more to the cost of 100LL.
    • That makes the cost of 100LL about the same as G100UL - GAME OVER.

 

I think you’re overestimating FBOs desire to sell a new fuel. It costs nothing to do nothing and until there’s a market demand and drop off of 100LL sales, nothing will be done.
It’s a chicken vs egg problem.

Let me know when substantial owners buy the STC. Right now that number is zero.

[1980Mooney]

26 minutes ago, ArtVandelay said:

I think you’re overestimating FBOs desire to sell a new fuel. It costs nothing to do nothing and until there’s a market demand and drop off of 100LL sales, nothing will be done.
It’s a chicken vs egg problem.

Let me know when substantial owners buy the STC. Right now that number is zero.

Perhaps my point was not clear.  I said once it is available in the supply chain then change will happen fast.  Some people talk like it will be a 10 year gradual adoption.  It will likely be more like a waterfall.  Yes I know that some people clung to their Sony Trinitrons long after higher priced digital flat screens appeared.  

Regarding sales of STC's I bet sales are brisk at Reid-Hillview County Airport (RHV) and San Martin Airport (E16).

[Hank]

9 hours ago, Pinecone said:

The only thing from preventing us from burning G100UL tomorrow is inertia.

Any refinery should be able to produce it now.  It is commonly made components that need to be mixed in the proper proportions.  Not rocket science.

The only thing from preventing us from burning G100UL is a lack of purchase orders. Place a qualified order with your favorite refinery for, say, 40,000 or 50,000 gallons and your FBO might buy it from you 8000 gallons at a time. That'll get GU100UL at one airport . . . .

It's always about the money. Refineries won't make it just to see if someone will buy it.

[A64Pilot]

What will drive the Gami fuel is LL being banned, I just hope its a State and not a National thing, but suspect it will be Federal.

Until LL is banned I don’t think the Gami fuel will have a market, when it is, then it will be the only available fuel.

I’m back to wanting Alcohol / water injection and 94UL

[A64Pilot]

4 hours ago, midlifeflyer said:

You mean like the Jet-A using, liquid cooled turbo, geared-prop in the newest DA-40S? Well, not particularly high RPM.

My only point is that the handwriting has been on the wall for a long time (we went from the Wright brothers to the moon in 60 years) but ignored. That's pretty much fact. Our interpretations might be different.

OK, one then, in one airframe.

Let’s see if it goes the way of the other Diesels

[Pinecone]

8 hours ago, jaylw314 said:

"Correlation does not mean causation" is only correct most of the time.  There has to be a point of sufficient evidence that the strength of correlation does mean causation for all intents and purposes.  That's how science works, otherwise, we'd still be banging the rocks together.   The tobacco industry had used that argument for decades.

In the case of lead causing health problems, that ship had sailed even long before it was used in gasoline with the immense amount of medical data throughout history.  Heck, even the makers of TEL called themselves  the "Ethyl Corp" because they figured "Lead" in their name would be bad PR.

The saying is based that correlation, by itself, does not mean causation.

It is one factor. 

 

[carusoam]

When do we suppose the first MSer pirep for the new fuel will be?

Price?

JPI downloads?

Savvy graphs?

 

Somebody wants to go first…

 

My first interview opportunity at the end of college… decades ago…

Company was called NL Chemical…

Their logo was a big NL on everything…

They did a great job hiding their name… National Lead!  

I don’t think anyone took the interview… even if it would be good practice…

Best regards,

-a-

[carusoam]

Now…

if we are looking for somebody to make lead weights for our elevators…

https://en.wikipedia.org/wiki/NL_Industries

NL industries is in TX…

 

At one time… early 1900s… NL started developing the lead free paints using TiO2…. And paid an artist to create  ‘the Dutch Boy’ for their logo…

hmmmmmm….   
 

Anyone recognize the connection?

NL = National Lead….

NL = NetherLands… the home of Dutch people

They have been trying to gently hide this connection to lead for a century!

 

Best regards,

-a-

[toto]

6 hours ago, carusoam said:

Somebody wants to go first…

I’d buy a tank of this stuff in a heartbeat if it were available locally. I’d love to get going.

100LL is eight bucks a gallon at my home drone, and it’s not doing us any favors

 

[Pinecone]

1 hour ago, toto said:

I’d buy a tank of this stuff in a heartbeat if it were available locally. I’d love to get going.

100LL is eight bucks a gallon at my home drone, and it’s not doing us any favors

 

Same here.  I am working on being the first Mooney to run it with the STC.

[N201MKTurbo]

1 hour ago, toto said:

I’d buy a tank of this stuff in a heartbeat if it were available locally. I’d love to get going.

100LL is eight bucks a gallon at my home drone, and it’s not doing us any favors

 

If you are expecting G100UL to be cheaper, I think you are mistaken. We will probably never know for sure. IF there are some airports that would have both, you would kind of know, but I have seen 100LL with a $2 difference at the same airport. In the beginning the production of G100LL will be fairly low quantities compared to 100LL, which will raise the price. 100LL currently has dedicated tank trucks. I realize they say they can be mixed, but will the refineries mix them? I doubt it. So they may need new tank trailers. As demand increases, will the refineries stop making 100LL and just make G100LL? I don't know. If you recall there was about 10 years with car gas where you could get both.

[aviatoreb]

18 hours ago, jaylw314 said:

"Correlation does not mean causation" is only correct most of the time.  There has to be a point of sufficient evidence that the strength of correlation does mean causation for all intents and purposes.  That's how science works, otherwise, we'd still be banging the rocks together.   

(Hypothetical story): Smoking causes yellow fingers.  Smoking causes cancer.  A wayward scientist unaware of smoking notices that people with yellow finger seem to be getting cancer.  Does a detailed study and finds a very high correlation between yellow fingers and cancer.  And then decides that yellow fingers cause cancer.  Then goes to the FDA and works to remove all yellow dies from food, from clothing and anything else that might yellow peoples fingers and thus cause cancer.

This is the problem of co-founders in causation inference and its one reason why even a very high level of correlation still might not imply that correlation implies causation.

[toto]

20 minutes ago, N201MKTurbo said:

2 hours ago, toto said:

 

If you are expecting G100UL to be cheaper, I think you are mistaken.

I’m certainly not expecting G100UL to be cheaper.

If you asked me a year ago, when 100LL was $5/gal, whether I would pay a $1-2 premium for UL, I’d probably be uncomfortable with that price. But we’ve crossed into new territory with high fuel prices, and now that I’m paying eight bucks, I’m ready to rip the bandaid off and get UL going. 

[Pinecone]

1 hour ago, N201MKTurbo said:

In the beginning the production of G100LL will be fairly low quantities compared to 100LL, which will raise the price. 100LL currently has dedicated tank trucks. I realize they say they can be mixed, but will the refineries mix them? I doubt it. So they may need new tank trailers. As demand increases, will the refineries stop making 100LL and just make G100LL? I don't know. If you recall there was about 10 years with car gas where you could get both.

Since all the components are normally made in refineries, it is just mixing things that are already available.  So I suspect that economies of scale will be less pronounced.

They can use the existing 100LL tanker trucks, or ANY tanker truck.  The reason that 100LL needs special trucks is due to the lead contamination.  It would take flushing and cleaning to use them for an unleaded fuel.

ANY refinery can make G100UL.  No separate processing or holding due to contamination with the lead.  So some refineries that are not making 100LL can make G100UL.

The margin on avgas is higher than auto gas.  So a refinery can make and offer at a nice price to break into a market they are not currently into.

[N201MKTurbo]

1 minute ago, Pinecone said:

Since all the components are normally made in refineries, it is just mixing things that are already available.  So I suspect that economies of scale will be less pronounced.

They can use the existing 100LL tanker trucks, or ANY tanker truck.  The reason that 100LL needs special trucks is due to the lead contamination.  It would take flushing and cleaning to use them for an unleaded fuel.

ANY refinery can make G100UL.  No separate processing or holding due to contamination with the lead.  So some refineries that are not making 100LL can make G100UL.

The margin on avgas is higher than auto gas.  So a refinery can make and offer at a nice price to break into a market they are not currently into.

My understanding is that the tank trucks are separate because of quality control not the lead. They can't guarantee the quality of the product if it was used in a tank where other products have been been stored, unless the tank is thoroughly cleaned, which would be another expense.

[jaylw314]

2 hours ago, aviatoreb said:

(Hypothetical story): Smoking causes yellow fingers.  Smoking causes cancer.  A wayward scientist unaware of smoking notices that people with yellow finger seem to be getting cancer.  Does a detailed study and finds a very high correlation between yellow fingers and cancer.  And then decides that yellow fingers cause cancer.  Then goes to the FDA and works to remove all yellow dies from food, from clothing and anything else that might yellow peoples fingers and thus cause cancer.

This is the problem of co-founders in causation inference and its one reason why even a very high level of correlation still might not imply that correlation implies causation.

The corollary to that is the (real) example of people who claim that physics, biology, pathophysiology, microbiology and western medicine are all lies because none of them have been 'proven', since all of the scientific process is based on correlation and not proof.  It's a bunk argument because the scientific process was never designed to 'prove' stuff, but it doesn't stop people who want to control you or sell you something from using that straw man argument.

What science is good a doing is disproving causation, and that is where your hypothetical example falls apart.  It is a trivial matter to design a study to disprove the hypothesis that yellow fingers cause cancer, and this is a common approach with confounding variables.

The only way science 'proves' stuff is if repetitive attempts to disprove a hypothesis fail.  This can be continued indefinitely, but practically, there at some point must be a conclusion that the mass of evidence, even if just correlation, is causation.  I agree that point is often misinterpreted as being inappropriately low by the lay media and the general public, but to continue to argue that correlation is not causation ad infinitum even with the accumulation of a large body of evidence is simply nihilism (which is again often the tool of people who want to stay in control or sell you something).

An analogy is the misconception that you require someone witness a murder to convict someone.  You can certainly convict someone of murder without direct evidence, but the quantity and quality of indirect evidence needs to be significantly larger.

Getting off topic into philosophy of science, sorry.

Edited October 18, 2022 by jaylw314

[ilovecornfields]

@jaylw314 and @aviatoreb - great discussion! I think this would have gone much better sitting around a table drinking adult beverages but I guess since you’re on opposite sides of the country this is as good as it gets.

@aviatoreb’s example is, of course, the classic example of confounding and was the one I was taught when I studied public health in grad school. Smoking causes yellow fingernails and smoking causes lung cancer, therefore people with yellow fingernails will have higher rates of lung cancer. 

Every study has the potential to give you misleading results and the most common causes are chance, bias and confounding. This is no secret, so someone who understands the field designs and interprets studies in a way to minimize these effects. 

As @jaylw314 pointed out and @aviatoreb is also keenly aware of - science doesn’t “prove” anything. That’s not know it works. But, when multiple well-designed studies consistently reject the null hypothesis (kind of like a double negative - “we didn’t find no difference between the groups”) then there is likely to be a true association between the variables.

Anyone who has studied chemistry learned the Bohr model of an atom (it’s the one that looks like a planet with a moon surrounding it). The Bohr model of an atom is wrong. It’s not the truth at all. Why do we use it then? Because it’s useful in predicting the energy levels of a hydrogen atom. All of our models, in fact, do not represent the “truth” but are representations of what is actually happen that are useful and allow us to do thing we couldn’t accomplish without them. Science gives us these models and these studies. Even the “yellow fingernails” study is useful because if you’re deciding who to screen for lung cancer and you decide to go with people with yellow fingernails then your screening will be way more cost effective regardless of the fact that the fingernails aren’t responsible for cancer.

At some point, we have to make choices based on incomplete information and this is where I think most of the problem lies. It’s really uncomfortable to make a choice, especially an important one, if you don’t know everything about it. Problem is, you will never know everything about something (and some of the things you may think you know are wrong). People deal with this discomfort in different ways - some just find someone they trust and listen to what that person says, some talk to others, some turn to scientific literature, others to religion, etc.. Then on top of this, we don’t interpret this information in a vacuum, but based on our knowledge, biases and different modes of cognitive processing. How confusing and frustrating!

Personally, I have to be comfortable making big decisions quickly with limited information because that’s my job- if you come in with stroke symptoms I have just a few minutes to decide whether you should get a drug that in some cases will help you and in others will kill you. And I have to make this decision with no way of “proving” that you even have a stroke and no way of “proving” that you won’t get a fatal intracranial hemorrhage if I do give you the medicine. But if I do nothing and wait too long, then you’re out of the window to get the medication and we lost that as a potential treatment option for your stroke. Sometimes I wish I had a job with easier decisions, like when to take the fries out of the crisper.

So, back to the topic of 100 LL it seems that most of us agree that:

Lead isn’t good for you.

We don’t want to pay more for gas.

We are mistrustful of how the transition to unleaded AVGAS will be implemented.

Can we go back to arguing about ROP vs LOP now?

[aviatoreb]

6 hours ago, jaylw314 said:

 since all of the scientific process is based on correlation

That is false.  I agree with your assertion after.

But I agree with what you said that "there is no proof" and my take - it is because scientific method is a concept where anyone can come up with an experiment at any time to debunk current dogma.  And it happens sometimes.   But experiments are counterfactuals - they are not correlations.  Correlations is not the thing at the heart of science not in the least, and I claim it is sort of a side show.  By the way, in mathematics we do have a thing which is truly a proof - based on a set of stated axioms, one can make theorems that must follow once a proof which is an air tight argument is made.  No where does a correlation computation show up in proving that the set of rational numbers are dense in the unit interval, for example.

There is an entire concept of causation inference based on interventions and counterfactuals.  I.e. experiments.

What if I create a test group of people whose fingers I deliberately paint yellow.  Does that result in an increased incident of cancer?

Beyond that assertion that all of science is correlation - which is false - there is something else that goes on with the concept of having a scientific method which allows a belief but a belief which can be challenged with experiments which either verify or challenge the assertion - it seems like that is a scientific method,  Not this thing called correlation which is a simple statistical computation.  I buy into that.  I believe until shown how to perform an experiment that shows me that my belief is flawed. 

And beyond that - the concept of knowing and believing is a philosophical question that is undeniably interesting but to my take not necessary for my life.   Do facts exits?  Does my concept of existence match your concept of existence?  If I see the color green and you say you see the color green are we experiencing the same thing?  If I drop a ball and every time I drop it, is that gravity or is that the hand of some deity playing mind games on me while I float in an immersion tank wearing silencing headphones and blinding eye mask piping an alternative reality into my head where I think I am dropping a ball and I think it falls due to gravity but gravity doesn't exist because its all a head game?

Here is some philosophy of causation

https://plato.stanford.edu/entries/causation-mani/

believe me a number of very smart scientists and philosophers have thought about what is causation over the thousands of years.  You can't just show up to the topic and assert things like all of science is correlation and expect that it has any meaning that someone has not already thought about.  There are a lot more significant other things going on in science besides simply correlates.

Edited October 18, 2022 by aviatoreb

[aviatoreb]

6 hours ago, jaylw314 said:

 

An analogy is the misconception that you require someone witness a murder to convict someone.  You can certainly convict someone of murder without direct evidence, but the quantity and quality of indirect evidence needs to be significantly larger.

I wonder if we mean the same thing when we say correlation.

In English, correlation means: "a mutual relationship or connection between two or more things."

https://www.merriam-webster.com/dictionary/correlation

which sure - means things are related.  Why not?

In math (statistics) it is a very specific computation with a formula that has many failings if over interpreted.  Besides confounding variables. missed variables.  It is also a linear response measurement tool.  And certainly things can be nonlinear too.

https://corporatefinanceinstitute.com/resources/knowledge/finance/correlation/

In your murder example - I don't think you mean you are computing a correlation coefficient.

I went to a fun invited workshop at Harvard of mathematicians and philosophers on what is causation really...

and I learned a number of good challenging concepts from the philosophers.  Here is one of my favorite.

Joe and Larry both pick up a rock and throw it at a window in an old broken factory.  But Joe is a bit closer.  His rock goes through the window first and the window breaks.  Larry's rock goes through the empty hole.  Who caused the window to break?  This challenges even counter factual.  Because if I could magically go and remove Joe from the experiment the window is still broken in the end.  So Larry caused the window to break.  But clearly also Joe caused the window to break.  Culpability might say Joe caused the window to break and Larry was lucky and did not cause the window to break so he shouldn't be challenged.  How about we run a correlation computation and call that science?

[aviatoreb]

3 hours ago, ilovecornfields said:@aviatoreb’s example is, of course, the classic example of confounding and was the one I was taught when I studied public health in grad school. Smoking causes yellow fingernails and smoking causes lung cancer, therefore people with yellow fingernails will have higher rates of lung cancer. 

As @jaylw314 pointed out and @aviatoreb is also keenly aware of - science doesn’t “prove” anything. That’s not know it works. But, when multiple well-designed studies consistently reject the null hypothesis (kind of like a double negative - “we didn’t find no difference between the groups”) then there is likely to be a true association between the variables.

 

..

At some point, we have to make choices based on incomplete information and this is where I think most of the problem lies. ...

Personally, I have to be comfortable making big decisions quickly with limited information...

But I again emphasize that a lot of science is not done statistically at all.  Even though I think in medical science it is more common.  And no one compute a correlation coefficient.  Look through a telescope and see a sunspot and refuse to say there is a solar flare going on because I forgot to compute the correlation for the event.  Run navier stokes equation in a computational fluid dynamics estimate for air flow over a wing and refuse to interpret the results since this is only one data point and so I can't compute a correlation.

Tell you what- your comfort with making important decisions with incomplete information, in crisis moments - as a medical professional - your my kinda guy and I only hope I would have you on call if I ever need you!!!

[ilovecornfields]

44 minutes ago, aviatoreb said:

That is false.  I agree with your assertion after.

But I agree with what you said that "there is no proof" and my take - it is because scientific method is a concept where anyone can come up with an experiment at any time to debunk current dogma.  And it happens sometimes.   But experiments are counterfactuals - they are not correlations.  Correlations is not the thing at the heart of science not in the least, and I claim it is sort of a side show.  By the way, in mathematics we do have a thing which is truly a proof - based on a set of stated axioms, one can make theorems that must follow once a proof which is an air tight argument is made.  No where does a correlation computation show up in proving that the set of rational numbers are dense in the unit interval, for example.

There is an entire concept of causation inference based on interventions and counterfactuals.  I.e. experiments.

What if I create a test group of people whose fingers I deliberately paint yellow.  Does that result in an increased incident of cancer?

Beyond that assertion that all of science is correlation - which is false - there is something else that goes on with the concept of having a scientific method which allows a belief but a belief which can be challenged with experiments which either verify or challenge the assertion - it seems like that is a scientific method,  Not this thing called correlation which is a simple statistical computation.  I buy into that.  I believe until shown how to perform an experiment that shows me that my belief is flawed. 

And beyond that - the concept of knowing and believing is a philosophical question that is undeniably interesting but to my take not necessary for my life.   Do facts exits?  Does my concept of existence match your concept of existence?  If I see the color green and you say you see the color green are we experiencing the same thing?  If I drop a ball and every time I drop it, is that gravity or is that the hand of some deity playing mind games on me while I float in an immersion tank wearing silencing headphones and blinding eye mask piping an alternative reality into my head where I think I am dropping a ball and I think it falls due to gravity but gravity doesn't exist because its all a head game?

Here is some philosophy of causation

https://plato.stanford.edu/entries/causation-mani/

believe me a number of very smart scientists and philosophers have thought about what is causation over the thousands of years.  You can't just show up to the topic and assert things like all of science is correlation and expect that it has any meaning that someone has not already thought about.  There are a lot more significant other things going on in science besides simply correlates.

As one of my learners recently said to me “that made my brain hurt!”

Very nicely written and excellent points.