Because I am using a Supercharged Ford V6 engine, I needed 39 PSI of fuel pressure at the fuel rail. I also wanted redundant fuel pumps for safety. To accommodate this requirement, a triangular shaped header tank is mounted behind the firewall on the passenger side. It fits between the structural angles behind the firewall, is about 2.5" thick, and holds about 1.5 gallons. Both fuel pumps (standard Ford T-Bird) are mounted in the header with a Tee connection between them. I have a switch for each pump mounted on the instrument panel along with a fuel pressure gauge. A float switch is mounted near the top of the tank, and turns on a red light on the instrument panel if the header is not full of fuel. A 1/4" aluminum tube is connected from the bottom of the tank to a quick drain on the bottom of the fuselage, because the header is the lowest point in the fuel system.
The two ports in each wing tank are in the standard position, one in front of the spar, and one at the extreme rear of the tank. To insure that air is not drawn into the line when one of the ports is uncovered in a low fuel situation, I borrowed an idea that Dick Marker uses in his BD-4. Both ports from the wing tank are plumbed to a solenoid valve (12 volt unit sold by J.C. Whitney for multiple tanks in trucks, ~ $30) mounted in the wing between the first and second ribs. The output from the valve is plumbed into the fuselage as normal, except there is a single 3/8" line. The solenoid valve is controlled by a mercury switch in the fuselage. When in a level or nose-down attitude, the front port is open and the rear port is closed, and vice versa in a nose-up attitude. I have a small LED on the instrument panel that lights when the valve is actuated and the rear port is open.
I really like the header tank idea. If fuel stops flowing to the header for any reason, the light will warn me. If power is reduced to "best range" speed immediately, there should be enough fuel in the header tank to stay in the air 10 to 15 minutes while looking for a comfortable place to set it down. Reduces the pucker factor a little.
Roger Mellema, Scott DeGaynor
I received a phone call from Scott DeGaynor the other day and he has yet another solution to the gas flow "problem" that some people have experienced with their BD's. He is installing a 5 gallon tank in the baggage area that is totally isolated from the all other fuel sources. This tank will have it's own gascolator and an electric pump. When he opens the valve to use this gas, a micro-switch will turn on the pump. The gas will be routed to a "T" fitting right at the carburetor. There will be a check valve in the line that comes from the regular gascolator so that this fuel will not be pumped back into the wing tanks. The reason for not connecting to the main tanks so that the aux tank can be refilled from them is the problem of fuel contamination. Scott will be extremely careful to make sure the fuel in the aux tank is clean.
I was particularly impressed by the number of fuel, or lack of fuel problems with the SD-4. I think this is caused by not only the flat wing, but more important is the fact that any system with a fuel pump that sucks as well as pushes will cause air, rather than fuel, to be drawn from any uncovered port that is connected to the system.
A totally gravity flow system (no pump) would let fuel run down any covered port. If there is a hump or high place in the fuel lines that is above the tank fuel level, gravity flow will not resume until the fuel level is raised above the hump, a slip will work to raise the fuel level to siphon fuel over the hump. A fuel valve with all 'on' porting used in conjunction with a fuel pump is more dangerous than a left or right valve because you double your chances of unporting a pickup.
The answer to me is to use a totally gravity flow system or gravity feed a separately vented header tank and install a push only type pump (turbine) in the header tank to feed the engine.
Enclosed is a sketch of the fuel system in my BD-4, note the header tank fuel warning lamp. It is a combination lighted rocker switch that lights when fuel head pressure at the header tank is less than 12 inches above the header. When this low pressure occurs a small diaphragm opens a small tank vent valve to atmosphere and turns on the warning lamp. The switch function opens the vent manually and is also a system check lamp. This feature lets me know when a tank is dry or a port becomes uncovered. I have had the light come on during initial acceleration with as much as half fuel, it goes out when the acceleration decreases, it also comes on longer with lower fuel level. The lamp will stay lit until level off when tanks are one fourth or lower. I have allowed myself sixty seconds of continuous'lamp on' operation then I will find a place to land before I use all of my one and one-half gallons in the header tank.
There is a filter sock on the fuel pump in the header
tank, screens on all six tank pickups, plus gascolator screen.
The fuel pump is from 1968 to 1976 Riviera or early Vega. It must be in the tank as designed. Output is sufficient at approximately 7 psi. Do not run the pump without fuel as it cools and lubes the pump. The header tank vents into empty first bay of wing. If you do not use the vent valve shutoff, fuel will run out of it when the tanks are full.
You may use this information if you like but like all homebuilt systems, each individual is on their own, I will not assume any responsibility.
Arguments Against a Header Tank
Re: your notes "more fuel flow words": It sounds like you did some careful work, and you came to similar conclusions as my experiments. One obvious thing you didn't mention, however; the apparent attitude of the airplane is a gravity vector, the vector sum of pitch angle and acceleration. At lift-off, the acceleration is substantial, and fuel starvation will occur at somewhat lower pitch angle than in steady-state climb. In my tests (done on a long runway) I could induce starvation with no pitch up on level run at full acceleration with minimum fuel (less than 5 gallons total). With 1/4 full tanks it was almost impossible to unport except by very steep climb, and then pitching over level would quickly restore power. I believe all the header tank solutions are causing more problems than they solve, or solving problems that aren't there. As you pointed out, moving the Y joint to a lower location is all that's really necessary. Scott DeGaynors rear-mounted header tank appears to be a move in the wrong direction - moving the tank aft and down is the wrong direction? He is ignoring the gravity vector.
I don't think Header Tanks are good because:
- They store fuel within the fuselage where the people are.
- They are complicated and add extra weight.
- People still get into trouble by taking-off with their main tanks empty or switched off and then getting just high enough to get hurt when the header runs dry.
Submitted 01/03/2002 in a letter, see also his fuel system article
Ram air "A" goes down to the header tank "B" and pressurizes it. When the fuel level "C" drops below the inlet line 'D" air goes back up the empty fill line to the empty tank either tank A or tank B. When you change tanks, the line from the header tank is pressurized and fuel can't flow from the fresh tank to the header tank. Excitement follows. Solution: Interconnect all the vent lines into a single system.