Operation Notice from Bob about Fuel Tanks on Systems with Fuel Pumps

[schu]

I guess I'll comment, hopefully it helps someone:

For years we were told to build to plans and NOT to modify the fuel system, and that it was dangerous to do so, but many of us knew that the bearhawk fuel system isn't adequate for every engine installation, specifically for any type of fuel injection that required a return.

Now we see this notice, but it's not really clear, and doesn't explain why. Someone asked if it's ANY fuel pump, to which Mark says yes, and given that Mark no doubt asked Bob, that is the official answer, but I think it might help to go through the different types of fuel systems and explain why design choices might make the engine quit.

Carb:

A carb doesn't need any fuel pressure, it just needs fuel in the float bowl, and the amount of fuel it needs is exactly whatever the engine uses. For this reason gravity works fine. The FAA says that a gravity fed fuel system must flow 150% of the fuel needed by the engine at a nose up attitude. So, you simply raise the nose, do a flow test and that's it.

Here are the gotchas: The FAA also states that when you use a fuel selector valve with the 'BOTH' position to tie the two tanks together, you also need to tie the vents together so that the pressure above and below the tanks stays the same. The bearhawk doesn't have this. In theory, with both tanks vented to atmosphere, they should flow the same, and this shouldn't be a problem, but there has been two reports of engine stoppage when one tank is very full and the other isn't when the pilot was using 'BOTH'. It's pretty obvious that in this condition for whatever reason, one tank is feeding the other and staving the engine. If I was running a carb I would have a cross vent, or I wouldn't use the 'BOTH' position. Even my Cessna 170 is placarded to use single tank operation at cruise, and it has a cross vent. All that said, many have been flying for years with no cross vent on both. Given that the bearhawk doesn't have hard points and exact plans to mount the fuel system, every single one is slightly different, so nobody knows why some work great, and others have had problems.

Bendix (returnless) fuel injection:

The Bendix system needs fuel pressure in order to work, so it requires the use of a fuel pump, and due to redundancy, two fuel pumps. Most people run an electric boost pump that can flow fuel when it's off, along with an engine driven pump. They operate at around 20-30psi. Because the fuel system doesn't have a return, the fuel pumps cycle fuel internally to regulate the pressure. What this means for the rest of the fuel system is that fuel delivery to the pump only really needs to be whatever GPH the engine is consuming. For this reason I believe the bearhawk fuel system works fine for returnless fuel systems. The flow of fuel down the door posts isn't really any different with returnless FI vs a carb.

Gotchas: You must manage the boost pump and turn it on when you need it. Also, like with a carb, it would be wise to use single tank or add the cross vent, but I don't think this is as big of a deal when fuel pumps are involved as flooded input port at the pump would almost certainly cause fuel to be pumped into the engine. The biggest issue with returnless FI (other than additional weight) is the fact that you have an engine driven pump which gets hot with fuel circulating in it which also gets hot. This can cause vapor lock if you aren't running a fuel that is very resistant to vapor lock (avgas). There are no certified airplanes with returnless fuel injection that have STC for mogas.

Return fuel injection (EFI and continental):

In the case of EFI, they need a bit of fuel pressure to work, and all of the ones I've seen pump the fuel through the injector rails, then to a regulator which bleeds the excess fuel back to the tank. This means that the fuel demand at the tank is whatever the pump consumes, which for some pumps can be quite high, like 80gph-100gph high. Of course, whatever the engine isn't using is put back into the tank through a duplex valve, but what if a tank becomes unported? Well, that pump is going to suck air pretty fast because it's consuming so much fuel. One way to deal with this is to try to never unport, so always use the fuller tank or only use one port (which you would want in the center), or use a header tank and return the fuel there. Personally, I would absolutely be using a header tank if I was going to run EFI. The continental runs at a lower pressure and doesn't return as much fuel, so it might work, and we have one member here that is doing it, but he put in 1/2" fuel line on the aft door post and also uses a L/R/Off valve.


Because I didn't want a carb due to the uneven fuel distribution, and also because I didn't want to completely redesign the fuel system, I'm going to run airflow performance with the factory fuel system and a valve that is L/R/Off. Landing will require selecting the fuller tank and turning on the boost pump. I don't expect any problems, but it is more complicated than a carb. The pro is that I'll be able to use LOP at low power cruise to cut down on my fuel costs, and I no longer have any issues with carb ice.


This not the official word from Mark or Bob, but hopefully it fills in the blanks for someone trying to work through what to do.

[schu]

Well, there is something new I didn't know. I assumed they cycled the fuel like a boost pump. Either way, if there isn't any flow, the fuel can get hot in there.

[zkelley2]

If by boost pump you mean the facet pumps that most people use, they also do not cycle fuel within themselves.

https://www.youtube.com/watch?v=Dtf66p8M11g

[svyolo]

I looked into a bunch of fuel system schematics and cutaways several months ago. All the diagrams I saw for Bendix style engine driven fuel pumps show an internal bypass to return excess fuel from the outlet back to the inlet.

The EFII and AFM boost pumps have to return fuel somewhere. The are constant displacement electric pumps. It is built into the external manifolds they come with in addition to a check valve to allow them to be bypassed. Same with the expensive certified boost pumps.

[PaulSA]

Thanks for the info Schu.

So, I have a standard fuel injected system with a back-up electric pump and no cross vent. Would you therefore recommend running LEFT/RIGHT in the cruise and fuller tank for T/O and Landing or BOTH for T/O and Landing?

[schu]

I'm going to run left/right and use the fuller tank because there are less ports to unport, because I'll certainly have much more fuel in one tank instead of both being low, and because I don't like how the bearhawk lacks the cross vent. I'll also probably use mogas in cruise from one side and land/takeoff with avgas from the other.

Battson runs BOTH, doesn't have a cross vent, and has hundreds of hours of trouble free operation.

I'm not sure what would work best for you. My solution requires much more fuel management, but I also plan on using fuel sensors with my EFIS as well as 30 minute alerts to check fuel to warn me if I'm being dumb. Battsons solution is dead simple, and very likely doesn't have issues with the lacking cross vent because as long as the pump inlet is flooded, it will pump fuel, even if there is some imbalance in the tanks.

Do you want to trade fuel management to have less unporting with a single, but fuller tank? That's up to you.

You could call Bob and talk to him too. I'm not the engineer. He may tell you you need a header tank....

[Bcone1381]

It seems prudent to me to have a limitation the says "Don't select BOTH if a significant fuel imbalance occurs". This sets the fuel system up to crossfeed from one tank to the other tank via the fuel selector. A crossfeed situation would not allow 100% fuel flow to the engine.

It also seems prudent to me to have little concern if both tanks quantities are essentially equal. Run it any way you desire as long as that limitation is honored and fuel is being supplied to the engine.

[Nev]

I was an earlier believer that a fuel pump could suck air rather fuel, but I'm not so certain now. On my aircraft, as long as any one of the four fuel tank ports is covered with fuel, then the fuel will make its way down the line and flood the pump inlet. I've tested this, and (on my aircraft) I'm satisfied that it works this way. I wrote a post about it HERE.

As a result, I always fly in the BOTH position.

It's my belief now that a number of fuel system issues that resulted in engine stoppages may be the result of flying out of balance (ball out to one side or the other) for extended periods of time. If a fuel selector is installed with only a L,R,Off position, it may remove the chance of unbalancing the fuel inadvertently, but increases the chance of unporting a tank if out of balance. On my own aircraft, if I select either the L or R tank, it becomes much easier to completely unport that tank if I'm flying uncoordinated, and this would lead to an engine stoppage. For that reason I would only have a fuel selector with a BOTH position.

It's been discussed at length before, but I wanted to add links to it on this thread for obvious reasons.

[Nev]

I should add that the sight gauges will appear to show a significant fuel imbalance between tanks if the ball is not centered. The imbalance will appear to be in the direction of the ball. It's very easy to mistake this when no imbalance occurs. If left for a while, an imbalance will develop.

Takeaway point - fly with the ball centered.

[Battson]

Always fly in the BOTH position.

Takeaway point - fly with the ball centered.

^ DO THIS.

I've been flying on nothing but BOTH for a decade and 1,000 hours now, in every possible fuel configuration you can imagine. Zero engine stoppages, zero problems. Not even a "splutter". I do occasionally use L or R to rebalance tanks, but that is very rare and I always set a timer, to alarm when the correct quantity of fuel has transferred.

I am firm believer in human error, making the fuel system or fuel management process any more complicated is inviting risk into your flying. Keep it simple, Sir.

[kestrel]

Just to offer an alternative view, not to say that everyone has to do it my way, I rarely fly on both. For some reason it is very hard for me to keep tanks balanced and it is an unstable situation. If one tank is heavier, it will tend to fly low and the imbalance is made worse. I _ALWAYS_ frequently glance at my fuel site tubes, even now that I have a fuel totalizer. I've heard of numerous stories where something unexpectedly cause fuel to be lost. Depending on a totalizer will almost always get you into trouble if this happens. Having "all your fuel in one basket" makes the situation even worse. The result is people not realizing that they have fuel problem until it is _ALL_ gone.

My aircraft is equipped with both engine driven and electric pumps. I have intentionally and accidentally (due to slips) starved the system of fuel. It always restarts immediately. I run both when low on both altitude and fuel.

Monitoring and switching tanks keeps me involved with my fuel situation and avoids surprises. "Set it and forget it" leads to problems not being caught early.

That's what works for me. I don't require others to do the same.

This post (not the entire thread) is an example:
https://www.supercub.org/forum/showt...l=1#post836492

[Nev]

I've heard of numerous stories where something unexpectedly cause fuel to be lost. Depending on a totalizer will almost always get you into trouble if this happens.

I guess a Totalizer can only measure the fuel that passes through the transducer, so in the case of a fuel leak or missing fuel cap then it can't measure that fuel. Hopefully now that there is a much greater awareness of the fuel cap issue (and the simple fix), lost fuel caps should no longer be part of the Bearhawk experience. I do the same and regularly glance as the sight gauges as a gross error check, bearing in mind of course that they only indicate correctly if the ball is centered - and it took a while to get my feet moving. When the ball is not centered, the sight gauges will indicate that one tank is very full and the other is very empty.

I run both when low on both altitude and fuel.

I'm with you on that (I run both all the time). There's unintended consequences to consider if the fuel selector is in any position other than BOTH in a low fuel state.


My own experience is that flying in balance (ball centered) is the largest factor on the fuel system.

[whee]

It’s fantastic that we get to build and fly our airplanes as we see fit. I enjoy the discussions relations to our choices and the camaraderie that comes from those discussions.

I’ve never had a plane with ‘Both’ and IMO it’s safer to have 5 gallons in one tank than 2.5 gallons in two tanks. Running a tank dry on a long cross country has always been part of my fuel management methods. I screwed up big time this past summer and wouldn’t have made it to an airport had I not ran a tank dry. Would have been better to not mess up the fuel planning in the first place but I’m glad the method I was taught in primary training saved my bacon.

https://www.avweb.com/features/avweb...fuel-tank-dry/

[rodsmith]

No instructor ever taught me this, my Uncle did, he routinely ran tanks dry in his Cherokee 6. I did this a few times in my Maule. Somewhere I have a large notebook full of printed Pelican Perch articles. So much common sense in those.

[kestrel]

There's unintended consequences to consider if the fuel selector is in any position other than BOTH in a low fuel state.

I run both in cases of low fuel and low altitude more because I don't necessarily know which way I might be slipping and so that I don't have to pay it as much attention. There are times I go Whee's way and deliberately run one dry. I've never seen a problem with a quick restart, but if circumstances suggest it, I will use one tank and take care with which way I let the ball go.

It’s fantastic that we get to build and fly our airplanes as we see fit. I enjoy the discussions relations to our choices and the camaraderie that comes from those discussions.

I’ve never had a plane with ‘Both’ and IMO it’s safer to have 5 gallons in one tank than 2.5 gallons in two tanks.

This is esp. true in an aircraft without both. This is our standard practice with our RV-4. It is what I do with the bearhawk if stretching fuel which is pretty rare given our 100 gal capacity.

Running a tank dry on a long cross country has always been part of my fuel management methods. I screwed up big time this past summer and wouldn’t have made it to an airport had I not ran a tank dry.

Why not? Same amount of fuel available.

I think the most important point is that if you are monitoring your fuel (...and your slip ball when low on fuel), then any method can work. Airplanes have crashed because someone ran a single tank dry at a very in opportune time. They have also crashed when they ran all their fuel out because they didn't notice that there was a problem when running on both. Neither problem will happen if you are actively monitoring fuel.

"Both" works well. Lots of planes don't have a both option and they are just as safe.

[Battson]

This "changing tanks regularly" approach to fuel management makes zero sense to me, in a high wing aircraft. It's nothing personal, just looking objectively at the risk level in each approach, there is a clear winner and loser. I also see no benefit to changing tanks, so what's the point in taking the extra risk.

That approach requires you to remember to do things, regularly, otherwise you get a surprise engine stoppage. Some people say it's a reminder to change tanks... It's a huge extra risk, not applicable in the other approach. Sure it works fine for lots of people and they have no problems, until you add in a lot of other workload or stresses - it's just a Swiss cheese risk model, an accident waiting to happen. It will work fine until the holes line up that one fateful day and humanity gets the better of you. We are all human and thus fallible, and people have been killed through fuel mismanagement so many times before. I don't think I am any better than they were, I am equally fallible.

At least on 'Both', it will keep running while there's fuel in there, no matter how bad everything else gets.

Also, "it always restarts immediately".... until the first time it doesn't, like when it happened to my good friend on short finals and the prop stopped spinning before he realised what happened. Fortunately he had enough energy to make the runway, just.

Also, if one is checking the fuel quantity all the time, what's the risk of having all the fuel "in one basket"?

If you are flying a Piper, then yeah - you have no choice.

It’s fantastic that we get to build and fly our airplanes as we see fit. I enjoy the discussions relations to our choices and the camaraderie that comes from those discussions.

One of the safety advantages of the high wing design is gravity feed to the fuel selector, I suggest pilots are in the business of managing and minimizing risk, and should have good reasons to expose ourselves, our families / friends, and our community to extra risk of harming ourselves or others. Especially when there is no clear reward for taking the risk.

[whee]

I agree with the points made and recognize that different methods can work and be just as safe as others.

To answer your question:

Why not? Same amount of fuel available.

Appropriately a gallon of usable fuel split between two tanks is within the margin of no usable fuel left at all. I’d argue that a gallon of usable in one tank is within those margins also but I’d rather it all be in one tank vs two. Maybe I would have made it on ‘both’ but based on my experience in my airplane I don’t think so. Either way, it was stupid and way too close.

[Nev]

Whee, I understand your logic with this. Out of curiosity, after running a tank dry, do you then fly with the ball slightly toward the dry tank to ensure the remaining fuel is fully ported ?

[whee]

No need to fly with the ball out of center. I ensured my fuel system flows enough fuel through the a single port to keep the pump flooded at all times.

[schu]

I think it's really interesting that Nev and I think kestrel mention that the fuel system doesn't draw air because there is enough fuel in the lines to keep a pump flooded. I'll assume both of you are using mechanical injection or a carb as an EFI system might draw enough fuel fast enough to pull a bubble in to the system only to cycle it and spit it back into the tank.

My airplane isn't flying yet, but I have yet another take on this. I'm using airflow performance injection, but I put in a duplex valve:

https://www.andair.co.uk/product/duplex-fuel-selector-fs22-d2/

I'm going to install a tee right after the mechanical fuel pump, crush a rivet in one end, drill a .020 hole in it, and return it back to the valve then to the tank. I have some testing to do to make sure I'm only cycling 5-8gph back to the tank at all fuel pressures, then I need to make sure that my fuel system will flood the inlet of the mechanical pump at 125% engine consumption plus that 5-8gph with nose high attitude, but if it all works out, I'll have a little bit of fuel cooling the pump all of the time, as well as a place for vapor to go. This combined with a shroud around the engine pump should allow me to run mogas on an injected lycoming. I'll probably install a fuel temp probe and do a LOT of testing, but given the cost of avgas these days, it made sense to have a plan to run mogas.

If it works, great! I'll be able to run LOP and burn cheap gas. If it doesn't work I'll pull the TEE and press on as I would choose injection over running mogas anyway.

As for left/right/both.... I'm with Whee and Kestrel, besides, I have a computer warning me to look at fuel every 15 minutes and duplex valves don't come in the both variety.

[JimParker256]

"Trust, but verify..." I use a combination of sight gauge readings and fuel totalizer readings to cross-check each other, and sanity-check that against expected fuel burn (simple @ 5 gph in cruise).

Call me paranoid if you will, but I haven't run out of fuel yet, and only once got into my personal 1-hour VFR fuel reserves when I was on a long cross-country with much heavier than forecast headwinds. Landed with 45 minutes fuel remaining that time...

My RANS S-6 Coyote II has no fuel selector - just a simple fuel shut-off valve. It has two 9-gallon wing tanks (with pretty accurate sight gauges) for a total of 17.5 gallons usable in the wings. The system gravity feeds from BOTH tanks simultaneously into a Y-fitting behind the cabin, with that single combined feed flowing into a relatively large 2.75 gallon header tank (2.5 gal usable). That gives the plane 20 gal of usable fuel. When I top it off, I set the totalizer to 20 gallons. If I partially fill the tanks, I just add 2 gallons to whatever the sight gauges total. The 1/2 gallon unusable in the wings is countered by the extra 1/2 gallon usable in the header, and this way the math is a lot simpler for me... Especially when distracted with other flight activities.

Being a very low dihedral airplane, it is a very common occurrence for this design to draw fuel from one tank for a while before the other one starts to drop at all. I've personally only seen a 4-gallon differential, but others have reported one sight gauge reading "empty" before the other one started draining, and even in those extreme cases, they experienced no engine issues. In fact, there are no known instances of fuel starvation in an S-6 that had fuel remaining in any tank. So that's helpful for my peace of mind.

But coming from several low-wing designs that had no "BOTH" selector, I will admit to being a bit "antsy" until I see the level of the second tank begin to drop as well, confirming that there is no blockage preventing fuel from flowing. If I get over 1 gallon out of "balance", I will intentionally fly a bit out of trim to raise the "full" tank higher so it starts feeding. Once I see that level start dropping, I know I've got fuel flowing from both tanks, and I go back to normal operation (flying in trim). And monitoring... Always monitoring... Don't ever want to be "that guy"...

[Nev]

This is a fantastic discussion. I like that others have their own methods, and they clearly understand the risks and have their own way of managing them.

In my mind it's very important that we come up with a safe way of managing fuel for people new to the Bearhawk. An SOP that doesn't require detailed systems knowledge.

My way of doing this during test flying was simply to keep the tanks at least half full at all times.
Keeping the ball centered at all times is the elephant in the room here. The Bearhawk has a very sensitive rudder, so until I was used to it the ball was all over the place - and so were my fuel tank quantities. Once I was able to fly consistently with the ball centered, I was then happy to run the fuel down to around 40 liters (10 Gallons) total (on BOTH).

This is one way. There are other ways to manage it, but it does need to be simple.
The overriding issue is the yaw sensitivity.

It's not a fuel system issue per se, it's the effect of yaw on the fuel system.

For a little light bedtime reading, more about it HERE and HERE.

[Battson]

For the benefit of new builders reading this thread, I want to reiterate that there have been several near misses and at least one fatality in Bearhawk aircraft due to apparent mismanagement of the fuel system and / or not flying in balance, apparently causing a tank to unport, per this safety message from Bob (original thread post).

Operational management of the Bearhawk fuel system (like all other aircraft fuel systems) is important to flight safety. The more complicated that activity becomes, the more likely accidents become. This has been proven time and time again in the aviation sector, at great cost.

I fully support the experimental process whereby people can do whatever they please. If people choose to modify the fuel system, that's got my support. We should be clear though, complexity comes with risk, and that tradeoff needs to be considered when making these choices.

Fuel injection systems with extra pumps and filters do complicate the fuel system, it's operation, and maintenance. They also offer a range of clear benefits in return for taking those risks, and it removes other clear risks such as carb ice. Implementing a fuel injection system with as little operational / maintenance complexity as possible, is a worthy goal. But it's not mandatory, and not as much fun as pure unbounded experimentation. We see the same choices presented when it comes time to operate the aircraft's fuel system.

Ultimately it's up to builders and pilots to decide.

I do think we need clear dialogue on this topic, as for years it's been mired in debate and uncertainty. This has definitely confused some builders who rely on this forum to help them make decisions.