What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Unique fuel system for injected engines

J

jim skala

Member
Vapor lock seems to be a leading cause of engine shutdowns for injected engines. How about a fuel system that CANNOT vapor lock, with an added feature that it can use most of the unusable fuel in a tank. Wouldn't that be nice?

Description: Use a nylon automotive screen at the tank pick-up. At entry to the fuselage from each tank, install a 3/8 inch fuel injection hose to a high-pressure electric fuel pump inlet. The outlet from each pump, 3/8 inch aluminum, routes to a TEE at the area of the fuel selector (which is no longer a fuel selector). 3/8 inch aluminum tubing routes forward to the firewall, connecting to the inlet of a LARGE size Andair Gascolator on the engine side. At the top side of the Gascolator is a 1/8 inch NPT tap that connects to a 1/4 inch aluminum tubing, immediately connected to a ball check valve (1/8 inch NPT fittings), connected to a fuel return line that terminates at the fuel selector valve, used to direct return fuel to either tank. The outlet of the Gascolator connects to the engine driven high pressure fuel pump. The outlet of that fuel pump connects to the fuel injector manifold. The ball check valve acts as an orifice to reduce fuel pressure from the electric pumps. A purist may want to install check valves around the fuel pumps that may allow fuel to easily be sucked in by the engine-driven fuel pump, but I found that it could suck fuel thru the engine-driven fuel pumps, so I did not do that. The positive sides of two fuel pumps are powered from two 10-amp fused (with resettable ATO fuses) circuits. The selection of a pump is from two redundant switches that connect the negative side of a fuel pump to turn it ON. These switches are the fuel selector. Note that it is possible to transfer fuel from one tank to the other. I installed a magnetic reed relay to sense the position of the return fuel selector, and this circuit activates a bright red LED on the panel immediately in front of the pilot when the selected pump is not returning to the same tank.
After I started flying with this system, I found a remakable, unintended feature: I could empty a tank down to less than two cups of remaining fuel. How this works: When the tank gets down to the last bit of fuel, the pump begins to suck a mixture of fuel and air, and the fuel pressure from the pump discharge begins to drop. The air/fuel mixture enters the Gascolator, and the air component immediately finds an exit and returns back to the tank. I found that the engine continued to run at full power with the injectors getting full pressure from the engine-driven fuel pump until the fuel pressure from the electric pump was below 10 PSI, at which time the fuel component of the air/fuel mixture was less than the fuel flow required by the engine. I immediately switched to the opposite tank and resumed full engine power.
 
There’s a whole lot that could be said here. Vapor lock is easy to mitigate with existing, long-proven, configurations. It’s just not that hard and the remedies don’t require added system complexity (e.g. avoid low velocity flows in hot areas like a filter or gascolator on the hot side of a FW. I’d suggest an open-minded FMEA of your whole system and especially those components that are unique to it. What are the consequences if one of these check valves sticks open or closed.

Fly safe, Sir

Moderator’s clarification…FMEA: Failure Modes and Effects Analysis, a rigorous analytical method used on the Space Shuttle Program and other Aerospace systems
 
Last edited by a moderator:
sorry, but I am not seeing the benefit to all of this. It appears that all you are doing is pressurizing the system all the way to the mech pump to help it overcome vapor lock. While this is a sound idea and will help in many cases, how is it any different than what the std boost pump is doing for the rest of us. Fuel recirc systems are great, especially for combating the heat soak scenarios like fuel stops (have them on both of my planes), but you are only recirculating the cold fuel in this arrangement. It is the fuel in the lines fwd of the FW that that gets hot and need recirc, as that is where all of the heat is. My recirc tap is directly at the servo.

Larry
 
Last edited:
As compared to a conventional RV system for constant flow injection, the described system is more prone to vapor lock if the electric pumps fail...for example, loss of electrical power. The dead pumps become a restriction, and as Scott notes, the gascolator is a big fuel heater.

The tee between tanks (rather than a selector) means that if you run one tank dry, failure of the opposite electric pump will allow the engine driven pump to suck air.

The system requires two red cubes or similar to run a totalizer. It won't be accurate if the return line is a mix of fuel and vapor.

These switches are the fuel selector. Note that it is possible to transfer fuel from one tank to the other.
Click to expand...

So with just one pump running, the system is transferring fuel to the dead pump side as well as supplying fuel forward.

There's more, but...
 
I gave it 24 hours. Time to look at all the responses...

"As compared to a conventional RV system for constant flow injection, the described system is more prone to vapor lock if the electric pumps fail...for example, loss of electrical power. The dead pumps become a restriction, and as Scott notes, the gascolator is a big fuel heater."
Yes, more prone to vapor lock 'if both electric pumps fail'. So you agree that it is not a single point failure. I have flown this system for over 1400 hours so far. I can turn off both pumps, and the engine-driven fuel pump takes over, sucking fuel from both tanks thru those "dead" pumps as long as there if fuel in both tanks. This actually happened to my hanger mate's plane (same fuel system) on the second flight, where his electrical system failed completely, and the place was flown about 50 miles back to the airport with both pumps dead. Also note that you would need a failure of the electrical system (both the main battery and the backup battery would need to go dead). There is no single point failure with a good battery backup (mine is a 20 amp-hour LiFePO4 battery, tested before every engine start-up). And how is the gascolator is a "big heater"? It is cooled by the fuel flowing thru it, and I doubt that it adds much heat to that fuel. And that fuel is under enough pressure to immediately collapse and vapor bubbles that could form.

"The tee between tanks (rather than a selector) means that if you run one tank dry, failure of the opposite electric pump will allow the engine driven pump to suck air."
True. So alternate between tanks, like a good pilot, and when you completely empty one tank, you should have maybe 3 gallons in the other tank, and THEN you would need the pump with the 3 gallons behind it to fail. I am more worried about bird strikes, and if I get down to 3 gallons, I should be landing anyway.

"The system requires two red cubes or similar to run a totalizer. It won't be accurate if the return line is a mix of fuel and vapor."
Could be true, but I use the Precision Airmotive EAGLE EMS, and it provides a very accurate fuel flow signal. Today, it showed 9 gallons remaining, and I found that I actually had 9.5 gallons remaining. Love it.

"There's more, but..."
I'm listening... shoot.

"These switches are the fuel selector. Note that it is possible to transfer fuel from one tank to the other. So with just one pump running, the system is transferring fuel to the dead pump side as well as supplying fuel forward."
No. The dead pump has a check valve that precludes fuel flowing backward thru it. The fuel return to tank is identical, regardless of which tank the fuel is being returned to. Transferring fuel when flying is slower than on the ground with engine off due to the fuel being used by the engine. Still, I find it useful to sometimes move a little fuel to the opposite tank in flight when a tank low fuel sensor activates. I did not mention: the two fuel selector switches are SPDT switches that connect the negative side of the pumps. As such, I could run both pumps at the same time, and that will produce an immediate high fuel pressure alarm, but the EAGLE system can handle it just fine. The wide band oxygen monitor changes very little during that transient from 25 to 45 PSI. I don't want to do that because I can only return fuel to one tank, and the big RED LED fuel alarm is annoying. However, in a fuel low emergency, I could (hypothetically) run both fuel pumps at the same time, returning to the higher tank to completely empty the low tank into the higher tank, and then run the one remaining tank to landing. This scenario would only be needed if fuel management were not properly addressed for the flight. Without this fuel system, that would not be possible, nor would it be possible to use the normally unusable fuel. I did not think of this when I installed this fuel system -- it was a complete surprise when I tested what would happen when a tank empties completely -- what would I see? I had to know. What other fuel system provides the ability to use literally all of the normally unusable fuel? I know of none. What happens when the fuel gets low and a slug of air comes into the fuel line? That should provide the pilot with some pucker power as the engine coughs a little from fuel starvation. It is also the reason that there is any unusable fuel.

"sorry, but I am not seeing the benefit to all of this. It appears that all you are doing is pressurizing the system all the way to the mech pump to help it overcome vapor lock. While this is a sound idea and will help in many cases, how is it any different than what the std boost pump is doing for the rest of us."
The std boost pump has no way to collapse vapor bubbles in the fuel from the tank to the inlet to the pump. And it fails completely when a slug of air gets to its inlet because it can only pump fuel. THAT is how it is different.

"Fuel recirc systems are great, especially for combating the heat soak scenarios like fuel stops (have them on both of my planes), but you are only recirculating the cold fuel in this arrangement. It is the fuel in the lines fwd of the FW that that gets hot and need recirc, as that is where all of the heat is. My recirc tap is directly at the servo."
The high pressurization from the discharge side of the fuel pumps collapse any vapor bubbles all the way to the injectors because the mech fuel pump is normally doing absolutely nothing -- it is there to add pressure when the inlet pressure is below 20 PSI. There is 18 inches of 3/8 inch injection tubing between wet fuel in the tank and the inlet of the fuel pumps. That is the ONLY place where I could ever see ANY fuel vapor, and when the fuel is very hot in the tank, I DO see vapor there. The recirc in my system stops at the gascolator. However, the fuel between the gascolator and the injectors does not get much time to heat up, maybe 10 seconds at most, so maybe one or two degrees at most? And at 20 PSI, the fuel is under enough pressure to preclude any fuel vapor. BTW, my EAGLE system does not use a servo, but multi-point injectors, one per cylinder, but that should make no difference. The whole point to present solid high pressure injection fuel at any temperature, and without any air bubbles.

"Vapor lock is easy to mitigate with existing, long-proven, configurations. It’s just not that hard and the remedies don’t require added system complexity (e.g. avoid low velocity flows in hot areas like a filter or gascolator on the hot side of a FW. I’d suggest an open-minded FMEA of your whole system and especially those components that are unique to it. What are the consequences if one of these check valves sticks open or closed."
I question that vapor lock is easy to mitigate. I have read many cases of panic due to vapor lock, so mitigation of vapor lock seems to still be an issue.
I was not going to let it happen to me. As for 'added system complexities', I don't see that at all. I see my system as fairly simple, and I see it as very effective at completely avoiding vapor lock. I have noticed that when I take off after letting the plane sit in the hot sun all afternoon, the usual 25 PSI at the gascolator is as low as 14 PSI. Obviously, the pump is sucking vapor bubbles that are collapsed before the pump discharge. The fuel pressure at the injectors drops to 20 PSI when the engine-driven fuel pump takes over, and the injectors see solid fuel, free of vapor bubbles. The same effect happens when a tank is close to empty, and slugs of air mix with fuel, and my system separates them at the inlet of the gascolator, returning that air to the selected tank. That is why I can use the normally unusable fuel down to less than a quart of fuel remaining in the tank. That alone is worth having this system. As for the last objection, "the consequences if one of these check valves sticks open or closed", the check valve in the return line at the gascolator is normally wide open. If it sticks closed, you get very high fuel pressure before you even start the engine, as it must be open to keep the fuel pressure under 35 PSI. If it sticks open, then you would have a problem only if the lose both pumps, so it is not a single point failure. The other check valves are on the discharge side of each fuel pump. If one of them were to stick open, fuel from the opposite pump would flow thru that pump to the opposite tank, causing a very low fuel pressure prior to starting the engine. AT that point, you replace the pump before you fly.

My biggest concern with this system is pilot error - returning fuel to the wrong tank or accidentally turning a pump off. To mitigate that concern, I have an obnoxious RED LED turn ON if I have the fuel returning to the opposite tank from the running pump, or if I have neither pump running, or if the ATO fuse on the running pump opens. That warning light depends on two magnetic reed relays that sense the position of the return fuel valve, and the circuit sees +12 volts on the negative side of either pump that is OFF. That is tested every time before I start the engine. Each pump is fed with its own fused circuit. And either switch can run either fuel pump, so a pump selector switch failure is covered.
 
DanH said:
The tee between tanks (rather than a selector) means that if you run one tank dry, failure of the opposite electric pump will allow the engine driven pump to suck air.
Click to expand...

This is a deal breaker for me. The whole reason we don't have a "both" selection on our fuel valve and just let it run until it gasps.
 
I’ve written many times about not having a “both” selection on the fuel selector in our RVs. It’s a disaster waiting to happen. Unport one of those tank pickups and the pump will only suck air.

BTW, one of the leading causes of accidents/engine stoppages in amateur-built aircraft is modification to the fuel system.

Vic

Vic
 
Last edited:
jim skala said:
Vapor lock seems to be a leading cause of engine shutdowns for injected engines.
Click to expand...

Do you have a reference for that assertion? From the NTSB data I’ve seen fuel exhaustion seems to be the leading cause of engine shutdowns. I’ve been flying fuel injected engines for decades and I’ve never experienced vapor lock nor do I know anyone personally that has either. IMO for RVs if you stick to the fuel system design per the plans there’s minimal risk meaning no changes in the basic design are required. YMMV……
 
vic syracuse said:
I’ve written many times about not having a “both” selection on the fuel selector in our RVs. It’s a disaster waiting to happen. Unport one of those tank pickups and the pump will only suck air.

BTW, one of the leading causes of accidents/engine stoppages in amateur-built aircraft is modification to the fuel system.

Vic

Vic
Click to expand...

I'm a proponent of a "better" fuel system than the ones we've had for decades - and I'm flying a pretty highly modified fuel system myself - but Vic is right. The quickest way I know of to put your airplane in the dirt is to play with the fuel system unless you are CERTAIN that you understand all the factors at play.
 
Very minor modification

My Rv-10 fuel system has been very slightly modified by adding a constant flow return line to the basic system.
It is made up of 1/4" fuel line and a separate fuel return selector allowing for selecting the return tank. The return line works well in conjunction with the purge valve return line which is a great feature on the Airflow Performance fuel injection system.
A single control Andair selector ($) with return feature would have been ok too.
No vapor lock in 12 years of flying, mostly on mogas.
 
Another 24 hours, and time to look again...

"The tee between tanks (rather than a selector) means that if you run one tank dry, failure of the opposite electric pump will allow the engine driven pump to suck air."

True, and this scenario seems to be the one that everyone fears. To arrive at that scenario requires the pilot to first run one tank dry, then requires a failed pump at the exact same time that the opposite tank is dry. Very unlikely. The TEE is not like having a fuel valve in 'both' because sucking fuel by the mech pump is not how the system normally operates.

My hanger mate built this system into his bird, but instead of the TEE, he used a 6-port Andair fuel selector. Lots of plumbing to overcome that one scenario. He retains the ability to burn the unusable fuel, which is the neatest feature of this fuel system, but he loses the ability to transfer fuel between tanks. As for me, I simply run both tanks down at about the same rate to avoid that feared scenario. The 3-gallon warning lets me know when a tank is low enough to be a concern, and if it is, I can always move a little fuel to balance the tanks, or I can burn the higher tank a little more.

'Vapor lock seems to be a leading cause of engine shutdowns for injected engines.' "Do you have a reference for that assertion?"

Google "airplane accidents from fuel vapor lock". There are many.

"I'm a proponent of a "better" fuel system than the ones we've had for decades"

I agree, and I flew my 'better' fuel system for 1400 hours before posting it. I am convinced that it is both better and as bullet proof as possible (as long as the pilot understands it). I could have used a pressure regulator at the return line from the gascolator, but the orifice approach is much less prone to fail. The 6-way fuel valve has complicated plumbing, each fitting a potential for a fuel leak. Such considerations went into the fuel system that I described in detail. I have several non-standard modifications that nobody commented on. The Vans fuel system uses a piece of fuel tubing with slots cut in it for a tank pickup. I use an automotive nylon filter bag at my pickup. It filters out just about everything except water -- but the gascolator filters out water.
 
Auburntsts said:
I’ve been flying fuel injected engines for decades and I’ve never experienced vapor lock nor do I know anyone personally that has either. IMO for RVs if you stick to the fuel system design per the plans there’s minimal risk meaning no changes in the basic design are required. YMMV……
Click to expand...

I had significant vapor lock in a RSA5 injected engine running AV Gas in the -6A we purchased, but there was the airflow performance filter and gascolator FWF. Long fuel line running around the engine between the metering device and the distribution block. Home brew plenum, no heat shield on the mechanical fuel pump.

After a long conversation with Airflow Performance, everything suggested was changed but the plenum. No gascolator, filter in the cabin, shorter, more direct hose routing, moved the fuel distribution block to the right side between 1&3 (for shorter hose routing), head shroud on the fuel pump. Now stock layout per plans. Much better.

As for the alternate fuel layout of the OP, glad it is working for you. Seems overly complicated. Probably close to 11,000 planes flying with a stock (or nearly so) working also. Fuel systems are not something i’d be willing to bet my life on if something I designed goes wrong.
 
jim skala said:
Google "airplane accidents from fuel vapor lock". There are many.
Click to expand...

Please. That’s not statistically relevant. I pulled up the last 10 years of NTSB accident data and I’m not seeing a lot attributed to vapor lock. I’m not saying vapor lock hasn’t been causal to accidents or downplaying the potential ramifications. What I am saying is the basic Vans design is sound and your assertion that something needs to be done isn’t supported by the available accident data. So we’re just going to have to agree to disagree.
 
Why ask?

Reading your lengthy comments and insistence on having a better fuel system makes me wonder why you are looking for validation at this point?
You claim to have 1400 hrs of trouble free flight time and from what I read, you are very much set on how well it performs, why ask for advice?
The problem is not so much your overly complicated system along with major flaws that have been pointed out by some very knowledgeable folks but rather something else that cannot be addressed in this forum.
I too do things that many don't approve but many do where in your case no one seems to agree that your system is sound.
 
Another 24 hours, and what do we have?

"Reading your lengthy comments and insistence on having a better fuel system makes me wonder why you are looking for validation at this point?
You claim to have 1400 hrs of trouble free flight time and from what I read, you are very much set on how well it performs, why ask for advice?"

I am not seeking advice or validation of any kind. I felt that perhaps the community might like a new perspective on a fuel system that can use the unusable fuel and could not vapor lock. I presented it, and there were none who could see the that perhaps the benefits outweighed the perceived risks. I answered all of the raised concerns of perceived risks. I think that perhaps the perceived risks stem from an incomplete understanding of this fuel system. Maybe the "not invented here" syndrome is in play. Does anyone know of another system that can completely empty fuel from a tank in flight? I know of none. Granted, when I installed this system, I did not know that it could do that.

Why I took this route over the Vans fuel system is that I saw big vapor lock risks in the standard design, with the facet pump pulling fuel a long distance from the tank thru a fuel selector valve, and dead-heading fuel, where it can vapor lock. Is that really such a great design? Just adding the fuel return improves the standard system a great deal, as some of the comments have noted.

I do not see this system as complicated or difficult to understand. I select a tank, and if I am returning to the opposite tank, I get an alarm. If I turn on both pumps, I get an alarm. If I turn off both pumps, I get an alarm. When I start pulling the unusable fuel from a tank, I see a gradual drop in fuel pressure at the gascolator for about 5 minutes before the fuel/air mixture cannot sustain full fuel flow to the engine -- not a sudden loss of power.

"The best system doesn't require any significant understanding."

I agree. How complicated is it to turn on a switch to select a tank and to turn a return valve to select a return path? Switching between tanks during flight is standard operating procedure. What else is needed to understand?
 
For what its worth--because we've done some non standard fuel system mods too, against the 'then' recommendations of the manufacturers.

Safety being paramount, we think out of the box when working on fuel systems. Whether is root mounted filters, or flow transducer locations, above everything else, the system has to flow fuel to keep the engine running. Without fuel to burn, the engine becomes quiet, and thats not a sound we want to hear on a flying plane.

I think we all can agree that fuel would rather be pushed under pressure, than to be pulled. In a suction system, like those with major components upstream from the pump, any place that isnt completely sealed will allow the disruption of flow. Lol take your shop vac, use it, then loosen the hose and see the difference. In the case of fuel and engines, this can be interrupted as vapor lock, but is actually a reduction in the suction pressure. Also consider, on a stock mechanical system, after the engine starts and the boost pump is turned off, the mechanical diaphram (not Romec) pump has to pull the fuel a long way from the tanks, and its dependent on a competely sealed system. This includes any component that has o'rings that seal shafts, housings, etc.

Conversely, pumps in the tanks, or the roots, push the full under pressure through the filters, lines, selector valve, and any other component. Same things apply, only under positive pressure. I agree that pumps in the roots, or even the tanks themselves solve some problems, but also bright to light other issues that have to be solved before a truely safe and reliable system can be fabricated. We messed with intank pumps in the past, but couldnt effectively and economically solve the swinging pickup issue. Then the pump switching issue meant something else for the pilot to do that was not necessarily normal operation. ( I understand that a company has potentially solved this with an Andair valve, but I have not seen it). The other issue I wasnt comfortable with was elevtrical power to the pump, submerged in a tank on an aircraft. Unlike a car or that stays on the ground, planes fly and that in itself can add some unknown factors.

So, while Jim's system works for him, it may not work for others. The "Bendix" type fuel system is tried and true, as is the SDS and EFii systems with their computer controls. I'd like to see pics of his system to see the components and their layout. Jim, if you would, please post the pics so we can see.

Tom
 
Vic made a good point on a teeing of tank feed lines or an equivalent "both" selector. It is a very important safety point. Aside from his example of why it isn't a good idea due to fuel pump sucking air, also keep in mind the FAA traditionally won't certify low-wing aircraft to select multi-tank feed. There is very good reason you hardly ever see this (with notable newer low-wing exceptions to be pointed out later). In order for a multi-tank fuel feed, most often the named "both" position on our small aircraft, there is a strict requirement that the vent air above the fuel in each tank must be interconnected in order for those tanks to be fed at the same time. So it takes two things to do the multi-tank feed boogie. A fuel feed line and an air vent interconnect line. There were many accidents and deaths prior to the 1940's when this airworthiness criteria was adopted. If the vent air is not interconnected and cross feed is allowed there is a chance the aircraft can suffer fuel starvation with one of available tanks still full or an alternate scenario where fuel is vented overboard emptying a tank. This was usually when one of the independent fuel tank vents would get plugged, usually creating a vacuum. Some of the smaller type aircraft actually had fuel tanks structurally collapse in flight due to the vacuum differential. This problem was easily fixed on high-wing aircraft because a vent interconnect line could easily traverse from side to side under the cabin headliner. Cessna found this out with the 120/140 series and the FAA mandated design change is why all high-wing Cessnas have been able to feed fuel on "both" ever since. This same air-to-air vent air interconnect is not as straightforward on low-wing aircraft due to the geometry. The vent interconnect line needs to maintain a higher level than the fuel throughout its path between the tanks which leads to some wild whoop-tee-doo plumbing with vent lines having to arch around the fuselage. Most low-wing manufacturers passed on that and went the simple foolproof method of having fuel selectors that only allow one tank selection at a time. Many of the later Part 23 low-wing aircraft like Rockwell Commanders, Cirrus, Columbia etc. were required to incorporate the vent interconnect and so can feed from "both". As an aside the Aviat Husky certified in the late 1980's only allows a "both" or "off" selection, go figure. Its certification was again part of the FAA's grandma-proofing desires in Part 23. Walking the flight line of homebuilts at OSH shows many aircraft with "both" selectors without evidence of venting interconnect. Some may go without fuel feed issues for their whole operational life, others might not.
 
Last edited:
Less than 24 hours this time...

The pictures will take some time because I will need the cowling off to take them. However, I can take pictures of the switches, the fuel pumps and the obnoxious RED LED warning light - just the stuff in the cabin. I just today flew the plane, and now I have to work Friday to make up for play time.
 
You must log in or register to reply here.
Back
Top