Uneven Fuel Feed From L and R tanks.

[IBob]

1 - A FAILED EXPERIMENT, OR FIX A PROBLEM AND MAKE A PROBLEM:

 

The Savannah has 35Litre Left and Right wing tanks (with sight glasses showing fuel level) feeding a 6Litre receiver tank behind the passenger seat.

The receiver tank has a vent pipe to return any air to the top of the L wing tank, ensuring that it remains flooded. The receiver tank is also fitted with a low level switch which makes when the level there drops (to approx 5Litres), lighting indicator/s on the instrument panel.

From the receiver tank, the fuel is passed to the engine via an isolator valve and the aux fuel pump etc in the usual fashion, with a return line via a small orifice (as recommended by Rotax) back to the R wing tank.

 

My Savannah has the 2 additional 35Litre 'long range' tanks, so my tanks are L Outer, L Inner, R Inner, R Outer. I have these individually valved at a manifold on the right hand side of the baggage compartment (where I can see them), essentially a copy of Mark Kyle's valving setup.

I am very happy with this arrangement.

I normally only use the Inner tanks, and we can ignore the outer tanks for the following:

 

I have always had the same uneven fuel take from my tanks, finishing any trip with the R tank 10 to 15Litres lower than the L tank. This did not bother me, as the fuel would take from the L tank once the R tank was down. And I was told that most aircraft have this to some degree. But I have been fiddling with it anyway, and one possibility was uneven fuel tank venting.

The standard Savannah tank vents are 4mm airline looped up from the fuel cap and down through the wing, with the open end in the airflow under the wing.

I recently saw a new Savannah where these vents were very neatly attached to the underside of the wing and directed forward, like mini-pitot tubes. I set out to copy this, and once I'd worked out how to get a tight bend into the air line without collapsing it, it came together very neatly.

 

Feeling pleased with my neat job, I fuelled up L and R evenly, and set out on a test flight.

My intention was to dip the tanks after landing to check again for uneven feed, since the sight glasses slosh around in flight and do not give an accurate picture. However, about half an hour out, I glanced at the sight glasses, and the R one seemed a long way down. Stranger still, the L one seemed a long way up.

So I turned round and went home, isolated and dipped the tanks after landing, and sure enough the R tank was way down, while the L tank had more fuel than when I set off. And the reason was clear enough once I got up to dip the tanks: despite the fuel caps feeling tight, there were fuel stains back across the L wing.

 

So my new vent pipes had pressurised the tanks, and since the L cap was not holding that pressure, fuel was feeding from the R to the L tank.

 

I have since returned my vents to straight down (then cut at an angle into wind) rather than facing forward, using the fastenings I had added to hold them there.

Thinking it through now, the smallest difference in tank pressure will result in some cross-feeding: running some rough numbers, a pressure difference of just 0.12PSI would be enough to entirely empty one tank into the other.

It seems to me it will be much easier to avoid this if vent pressures are reduced.

 

[Downunder]

Run the fuel return to the "receiver/surge" tank and only open the valve on the tank you want to draw fuel from. 

 

Edit: If the return needs a vented tank and the receiver tank is not vented then this may not work I guess.....?

Edited November 29, 2020 by Downunder

[IBob]

That's certainly a possibility, Downunder, though I would prefer to be able to fly (locally) with the 2 inner tanks valved on, so avoiding the need for monitoring and valve changes (and relying on the receiver low level switch, which itself is not failsafe).

 

Certainly if I was building again I would run the fuel return to the fuel valve manifold (it's closer than the receiver tank), rather than to the R Inner, though this is only an issue for me when running on the outer tanks with inner tanks fairly full.

Otherwise, note that I normally land with the R Inner tank low despite the fact the return line goes to that tank.

[Old Koreelah]

26 minutes ago, Downunder said:

...If the return needs a vented tank and the receiver tank is not vented then this may not work I guess.....?

I learned the hard way that all tanks need to be vented. 

[IBob]

The Savannah receiver tank is normally vented to a point high on the L Inner wing tank. The principal purpose of this is to ensure the receiver is flooded, avoiding air in the receiver which can result in false 'low fuel' indications from the switch in the top of the receiver.

Edited November 30, 2020 by IBob

[Downunder]

3 hours ago, IBob said:

Otherwise, note that I normally land with the R Inner tank low despite the fact the return line goes to that tank.

Perhaps "one up" your left wing is a touch low and it cross feeds?

 

I've not seen an aircraft (of our type) flow evenly from both tanks. Even the expensive Euro types.

Whether it's flying out of balance or a wing low.....or poor venting.

 

I do "manage" my fuel from both my tanks, each with it's own valve.

This helps balance the aircraft with more in the right tank, one up.

I also switch tanks at certain gauge points to better estimate fuel quantity remaining.

For example, if I switch from the left tank at X on the gauge and then run the right tank down to X on its gauge, I can cross reference that (known amount) with the flow meter reading.

I feel happier being in control of my fuel than just sitting there "wondering" what's going on.

 

As for the return line, I feel high wings with a good flowing fuel sytstem don't need one. (No choke points to the pump under 8mm dia and good flowing filters.) There is already head pressure assisting flow and working against vapour lock.

I can understand why Rotax brought out the SB. To cover their asses and all bases.

A low wing (with no electric "push" pump) on a hot and high day is a recipe for vapour lock without a return line.

If people are happier fitting them that's ok......

 

Edited November 30, 2020 by Downunder

[Nobody]

The way that the cessna wing tanks work(at least on the 172) is that there is a vent that runs form one tank to the other and one tank is vented to the outside. This means that the tanks are always at the same air pressure ensuring that the levels in the two tanks are fairly equal. 

 

It is very hard to get two air vents to be at exactly the same pressure as you have found out. Any difference in air pressure will cause the fuel to flow to one side.

[IBob]

Downunder, you read my mind.

 

2 - AM I FLYING SOLO LH DOWN A BIT?

 

The outer long range tanks came in handy for checking this one: I worked out that my offset weight in the driving seat was equivalent to 15Litres of fuel offset in the R Outer tank. So I tipped 15Litres of fuel into the R Outer tank (L Outer obviously empty) and went for a fly.

It made no difference at all: I'm still landing with my R Inner tank 10 to 15Litresd below my L Inner tank.

[IBob]

4 minutes ago, Nobody said:

The way that the cessna wing tanks work(at least on the 172) is that there is a vent that runs form one tank to the other and one tank is vented to the outside. This means that the tanks are always at the same air pressure ensuring that the levels in the two tanks are fairly equal. 

 

It is very hard to get two air vents to be at exactly the same pressure as you have found out. Any difference in air pressure will cause the fuel to flow to one side.

Nobody, thank you, I wasn't aware of that, and it's given me something to think on.

Putting on my fix a problem and make a problem hat (something I seem to have a talent for): with the 172 arrangement you describe you'd need to be very sure that single vent could not be blocked, or you would lose all fuel feed....

[Nobody]

3 minutes ago, IBob said:

Nobody, thank you, I wasn't aware of that, and it's given me something to think on.

Putting on my fix a problem and make a problem hat (something I seem to have a talent for): with the 172 arrangement you describe you'd need to be very sure that single vent could not be blocked, or you would lose all fuel feed....

I have just looked at the diagram again and I was wrong in my previous post. While there is an interconnecting line and one vent, one of the caps is also vented too. The Cessna diagram is in the POH page 7-21.

https://takewingaviation.com/wp-content/uploads/2017/02/N739EF_172N_POH.pdf

 

It does mean that both the vent and the opposite cap would need to get blocked to cause fuel starvation. The key difference compared to your fuel vent setup is that the tanks are cross vented to put them both at the same pressure.

[IBob]

Nobody, thanks again.

I believe I have seen the underwing vent on a 185: from memory approx 3/8 pipe bent forward like a stubby pitot.

Interesting to note it has a check valve: I imagine this would be a simple swing valve.

 

Yes, given the large effects of very small pressure changes, the cross-venting makes every sense.

[RFguy]

this topic is EXCELLENT

gives me some things to look for in the uneven fuel usage (if on BOTH) in the club's Brumby 610. (912ULS) .

Edited December 1, 2020 by RFguy

[Old Koreelah]

All my engine stumbles and stoppages have been caused by venting issues. Getting the shape and location of the vent correct is crucial to ensuring reliable fuel supply to the engine. Even a minor change in airflow past the vent can make a major difference to air pressure in the tank, dictating whether it flows or not.

[facthunter]

Skid ball to one side will give the same effect as wing low. It's the difference in pressure that is the cause with the vents. Making pressure common may mean it can cross feed and vent when tanks are full or aircraft parked across a slope.. SELECTING only ONE is the only way to absolutely have control but the Problem is forgetting to change tanks at required times. I also like to have more than one tank selected at take off but if one sucks air you starve for fuel. Pumps in each tank with NR valves are the only way to use every drop. Nev

[IBob]

Some really good input, and much to think about here for me. Thanks to all.

 

In many ways the Savannah is an easy case, being high winged.

And the 6litre receiver behind the seat, with low level switch in the top, is an excellent part of the setup: provided the switch and indicators are working, the pilot is alerted when down to 5litres of fuel in the receiver. (And the one proviso here is that the test button on the panel tests only the indicators, it cannot test the switch. So if intending to use the indicator to fly a tank dry, I first test the low level setup on the ground by valving off all fuel to the receiver, and looking for the indicators to come on during initial runup and taxiiing.)