Level switches in the fuel system for alarms - Yes, No, Why?

Discussion in 'Firewall Forward / Props / Fuel system' started by wsimpso1, Dec 14, 2018.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Dec 15, 2018 #21

    lr27

    lr27

    lr27

    Well-Known Member

    Joined:
    Nov 3, 2007
    Messages:
    3,216
    Likes Received:
    463
    The strain gauge installation I figured out was pretty easy. However, I didn't have to worry about the display device, as the buoy already had a to d stuff, a little computer, etc. I didn't realize this had to go on the Dynon, was thinking of driving some LED's or a little volt meter. I'd put the amplifier right on top of or next to the strain gauge, so any long wires would be carrying a reasonable voltage. Could possibly put a capacitor in the output to slow down the response a few seconds to keep sloshing from setting it off all the time.

    I did point out that if your time is worth much, the commercially available, pre-made unit is better. The parts for the gadget I have in mind would be very inexpensive. I don't know how much they cost from the usual sources, but you can get a full bridge strain gauge for $4 on eBay. My guess is that the whole thing might be around $20. You might not even have to remove the tank from the airplane.

    I'm not saying that a strain gauge is necessarily the best way, but I'm pretty sure it would be a viable way to do this.
     
    wsimpso1 likes this.
  2. Dec 15, 2018 #22

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,546
    Likes Received:
    2,572
    Location:
    Calgary, Alberta, Canada
    Our EFI system outputs a digital FF pulse train emulating the old type Red Cube mechanical transducers. Accurate within 1% once calibrated correctly. This should make gauges redundant as the Dynon totalizer will show fuel remaining if you enter the correct figure when you fill the tanks.

    I never really trust or look at gauges too much. I know the fuel burn and how much is fuel is on board. A little mental math and I can figure 4 hours plus an hour reserve or whatever the case may be. I usually don't want to sit there for more than 3.5 before stretching, so everything is good.

    With your setup, I think you just need a red light and aural alert you can silence once the mains drop below 3-5 gallons or so. Ditto on the header tank. I recommend never running a tank really low as you may find the next tank you select won't feed and then you're options may be limited, depending on terrain and how far away airports are. If you never push it, you never have a pucker factor on fuel. Never worried about fuel in 15 years of flying my RV but it has a much simpler fuel system than you do. I wouldn't over complicate the system. Aural alerts are way better than lights which can be missed with bright sunlight on the panel or when task saturated.
     
    wsimpso1 likes this.
  3. Dec 15, 2018 #23

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,028
    Location:
    Mojave, Ca
    Time for a reset... I got too wrapped up in details.

    To your concerns:

    1. Dont worry about the low level of the wing tank too much - you have header to keep the engine running and the Facet type transfer pumps dont die when run dry anyway. If you really want a reminder, a single float switch will tell you when the wing is empty. Easy solution.

    2. As mentioned by Ross, if you run the SDS system, the ECU has a built in FF output. Easy solution.

    3. An appropriately baffled header tank will have a pretty stable fuel level. Another simple float switch will tell you if the fuel level is really dropping below "normal". If it does, thats a positive indication of a wing feed issue long before it becomes a legitimate emergency. Easy solution.
     
  4. Dec 15, 2018 #24

    proppastie

    proppastie

    proppastie

    Well-Known Member

    Joined:
    Feb 19, 2012
    Messages:
    3,503
    Likes Received:
    1,005
    Location:
    NJ
    A little more difficult if you do not fill up but if one most always fills up then it is a few button pushes to bring the number back up to 54 gal. And yes I have forgotten but I also have accurate fuel gauges which might seem overkill but I set my power by gal/hr which is real nice and easy.
     
  5. Dec 15, 2018 #25

    Marc Zeitlin

    Marc Zeitlin

    Marc Zeitlin

    Well-Known Member

    Joined:
    Dec 11, 2015
    Messages:
    424
    Likes Received:
    571
    Location:
    Tehachapi, CA
    I built AUX tank for a COZY III that was going to do an around the world flight. I had installed capacitance level senders in the two strake tanks a couple years prior. We put capacitance senders in the AUX tank as well. Not completely analogous, since your tank should mostly be full, but I can see no reason NOT to have a level sender in the header tank. Sure, it'll be full 99% of the time, but if it isn't, you want to see the level. Rather than cobble together some system that only gives you one or two points of fuel level, you can have ALL levels indicated. Set the yellow alarm for 10% under full, which will wake you up and let you know that something's going on while the tank is still almost full, and red for 75% full, which will be the indication to get the hell on the ground.

    The EMS will yell at you and there's no additional work to figure anything out for an install.

    Between the level senders and the totalizer (not to mention your watch), you really have to TRY to run out of fuel.

    Don't complicate things unnecessarily. You've got a fuel tank - give yourself a level reading. You want to put a low level sender somewhere - put it in your sump for low oil level :).
     
    wsimpso1 likes this.
  6. Dec 15, 2018 #26

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,625
    Likes Received:
    2,866
    Location:
    Saline Michigan
  7. Dec 15, 2018 #27

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,625
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Good to know the Facet pumps won't give up quickly if the wing tank goes empty.

     
  8. Dec 15, 2018 #28

    Mad MAC

    Mad MAC

    Mad MAC

    Well-Known Member

    Joined:
    Dec 9, 2004
    Messages:
    388
    Likes Received:
    151
    Location:
    Hamilton New Zealand
    Accurate fuel gauges should always be the starting point for aircraft fuel systems. Locally there was a student pilot on a cross country that came close to a "forced landing" until talked out of it by an instructor over the radio because the fuel flow gauge was telling him (by Voice Annunciator) that the aircraft was almost out of fuel. Turns out the on the previous flight someone had set the fuel totalizer quantity.
     
    proppastie likes this.
  9. Dec 15, 2018 #29

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,028
    Location:
    Mojave, Ca
    "Accuracy" is not even required by regulation. All that is required is a positive indication the tank is empty. And that can have a very wide interpretation- as displayed here.
     
    proppastie likes this.
  10. Dec 15, 2018 #30

    proppastie

    proppastie

    proppastie

    Well-Known Member

    Joined:
    Feb 19, 2012
    Messages:
    3,503
    Likes Received:
    1,005
    Location:
    NJ
    correct, but accuracy is a nice feature to have, especially near the empty part of the tank.
     
  11. Dec 15, 2018 #31

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,028
    Location:
    Mojave, Ca
    In this particular case I think it depends on the use case of the header. If its large enough to be used as a practical "last fuel tank" rather than simply as an accumulator/snubber tank, then yes, I'd like some level of Qty indication beyond "not full". In other words, is normal ops to deplete the left and right tanks, fly a bit more and still land with reserve? Or is normal ops to land with the reserve in a wing tank, depleting wing fuel only in an emergency?
     
    wsimpso1 likes this.
  12. Dec 15, 2018 #32

    Marc Zeitlin

    Marc Zeitlin

    Marc Zeitlin

    Well-Known Member

    Joined:
    Dec 11, 2015
    Messages:
    424
    Likes Received:
    571
    Location:
    Tehachapi, CA
    Missed this the first time through.

    I have a Dynon, with fuel sight gages (COZY MKIV), fuel level senders, and a totalizer/FF gauge. While I agree that "IF you enter the correct figure when you add fuel", about 1% - 3% of the time I get gas I forget to put the totals in the system. Yeah, we can argue about checklists, but I'm still a human being...

    So next flight, the level gauges and the totalizer don't match. Since there's no manual labor involved in the gauges, but there is in the totalizer, I always assume that I screwed up and the gauges are correct and the totalizer is wrong. If I didn't have the gauges, I wouldn't actually have a clue how much fuel I had on board (if I have baggage in front of the sight gauges in the back seat). After calibration, my level gauges are good to within about 1/2 gallon at all times.

    The last reason to have level gauges along with the totalizer is to indicate leaks. If you have a leak from a tank, the totalizer won't know it. Start with 40 gallons on board, burn 20 of them, and you'll have 20 gallons left on the totalizer. But if you've leaked out 15 gallons, the totalizer won't know, but the level gauges will. If you think that you'll never have a leak, then this won't be a concern. But how can you possibly guarantee this?

    So I think that _IF_ you're only going to have one type of fuel quantity indicator, it should be a level gauge, with yellow and red alarms set on the EMS. The totalizer is a great backup for fuel quantity, and the primary indicator of range and economy.

    My $0.02.
     
  13. Dec 15, 2018 #33

    Dan Thomas

    Dan Thomas

    Dan Thomas

    Well-Known Member

    Joined:
    Sep 18, 2008
    Messages:
    4,706
    Likes Received:
    1,939
    Here's a digital version: https://www.electroschematics.com/5979/door-timer-with-alarm/

    It could be done with resistors, capcitors and a transistor or two, but it would be bigger and I'd have to experiment to get the values. I never got to university to learn the math to figure the values. It would involve a high-value resistor feeding a capacitor-transistor base connection and another resistor connected from there to ground. The capacitor's negative side would also be grounded. The voltage on the transistor base would rise slowly as the capacitor filled until the base received enough current to fire. Might need a zener diode in there to act as a holdback. The other resistor would drain the capacitor when the switch opened. It would all be rather cumbersome and inaccurate, and the capacitor would need time to drain between switch closings.

    Edit: what was I thinking? Just buy something like this: https://www.amazon.ca/Timer-Delay-Relay-Hours-Cycling/dp/B00PD65UGA
     
    Last edited: Dec 15, 2018
  14. Dec 15, 2018 #34

    Marc Zeitlin

    Marc Zeitlin

    Marc Zeitlin

    Well-Known Member

    Joined:
    Dec 11, 2015
    Messages:
    424
    Likes Received:
    571
    Location:
    Tehachapi, CA
    First, let's talk about the "new" Part 23, which we (as EAB aircraft) are not required to follow, but it's generally good practice. 14 CFR 23.2340(a) says:

    (4) Provide the flightcrew with a means to determine the total useable fuel available...

    While accuracy is not defined, if it's not reasonably accurate, then the flightcrew will have no idea as to the total usable fuel. So the notion that the only requirement currently is for positive indication of "empty" is not supported by the FAR. Of course, in an EAB aircraft, there's no requirement for fuel level indication at all, so there's that...

    With respect to the old FAR and the old wives tale of "only need to know when it's empty", here's what the previous version of Part 23 had to say. 14 CFR 23.1337(b):

    (b) Fuel quantity indicator. There must be a means to indicate to the flightcrew members the quantity of usable fuel in each tank during flight. An indicator calibrated in appropriate units and clearly marked to indicate those units must be used...

    Same comments apply as for the new version. There's a good discussion of this question here, with explanations for what the problems with the regs are and why the OWT is incorrect:

    https://www.av8n.com/fly/fuel-gauges.htm

    particularly section 3, which discusses the regs.

    Again, we aren't held to the requirements of Part 23, but there's a reason they were written...
     
    wsimpso1 likes this.
  15. Dec 15, 2018 #35

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,028
    Location:
    Mojave, Ca
    Once again, we're getting into semantics. The salient point is that it is very difficult to show an accurate fuel level in a long, skinny tank like is found on most light aircraft. The typical float type resistance senders are mechanically "topped out" for the top 1/3 to top half of the actual fuel QTY. I can fly for well over an hour in the RV and the tanks both read "FULL" the whole time. By the time my needles come off full, I've burned a significant portion of my usable fuel. Such a system is not a particularily useful "...means to determine total usable fuel available..." Of course this is not a new problem, and thats why the only realistic measure of accuracy is on or near the bottom. This is why I always put eyeballs on the actual fuel level in the filler neck of each tank before I take off, and always do the quick mental calculation for "time available" aloft. I also have forgotten to reset the totalizer, but it is rare. I have come to rely on the totalizer as my primary (minute to minute) assessment of remaining fuel. But I have never been comfortable with fuel qty guages at face value unless they are near empty.

    The botom line is that there are many ways to meet Part 23 (or the intent of) without a direct reading indicator in each tank. I can think of several current use military and airliners that lack direct reading indications for large portions of fuel. The design of the fuel system and the expected use case should drive the most effective human/machine interface. If thats direct reading fuel qty in every storage vessel, or a totalizer and a peek in the tanks at the begining of each leg, so be it. Whatever contributes most effectively to the pilot's management of the flight.
     
  16. Dec 15, 2018 #36

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,546
    Likes Received:
    2,572
    Location:
    Calgary, Alberta, Canada
    You mentioned it, but fuel on board should be visually inspected pre-flight and totalizer amount should be one of the things on the startup checklist in this case. I think we should always have a rough idea of how much flight time is remaining for the fuel on board, just in our heads (assuming no leaks).

    Fuel leaks is a valid point for sure. I can think of two forced landings which happened due to fuel being pumped overboard through a fitting leak but in both cases, the pilots noted but didn't believe the fuel level gauges which didn't jibe with FF indications vs. time aloft.

    I think with a larger header tank of 5-6 gallons, a level switch triggered at the 10% down level may still be all that's required there and at the 5 gallon remaining points on each main. This would give you a minimum of about 10 gallons remaining and hopefully 15 if you didn't run the other main below 5 to begin with. I agree a good level reading might allow you to evaluate a leak situation sooner than a low level switch. Regs might require it too. I haven't looked at the specific wording in a long time. Lots of things come into play with terrain you're over (like water, mountains). If you have a leak or can't feed from one main, it could put you in a pickle no matter what.

    All useful discussion here I think.
     
    wsimpso1 likes this.
  17. Dec 15, 2018 #37

    Dan Thomas

    Dan Thomas

    Dan Thomas

    Well-Known Member

    Joined:
    Sep 18, 2008
    Messages:
    4,706
    Likes Received:
    1,939

    But you are held to the requirements of FAR 91.205, which has this to say:

    (a) General. Except as provided in paragraphs (c)(3) and (e) of this section, no person may operate a powered civil aircraft with a standard category U.S. airworthiness certificate in any operation described in paragraphs (b) through (f) of this section unless that aircraft contains the instruments and equipment specified in those paragraphs (or FAA-approved equivalents) for that type of operation, and those instruments and items of equipment are in operable condition.

    (9) Fuel gauge indicating the quantity of fuel in each tank.


    From:
    https://www.ecfr.gov/cgi-bin/text-idx?SID=fb9fdbae6f6e22c8678f5c3e5e42c4ad&mc=true&node=se14.2.91_1205&rgn=div8

    In Canada, we have it worded this way:

    605.14 No person shall conduct a take-off in a power-driven aircraft for the purpose of day VFR flight unless it is equipped with

    (j) a means for the flight crew, when seated at the flight controls to determine

    (i) the fuel quantity in each main fuel tank...


    So, a dipstick does not qualifyas a fuel gauge.

    And yes, that OWT about the gauge only needing to indicate empty when the tank is empty is a stubborn fable spread by people who have never actually read the regulations. And if they read FAR 23 in any detail they'd see that "Empty" means "unusable fuel" level, not totally empty. Unusable fuel is that fuel that will not flow out of the tank in some normal attitude such as a Vx climb or full-flap, power-off glide. The unusable fuel, for certificated aircraft, is the difference between the tank volume as posted next to the fuel filler and the amount posted next to the selector or shut-off valve. Those numbers can be found in the POH or TCDS, and some TCDS post the unusable fuel clearly, such as this example from the 172 TCDS:

    Serial Nos. 17261578, 17261445, 17265685 through 17274009
    The certificated empty weight and corresponding center of gravity location must include unusable
    fuel of 24 lbs. at (+46) through 172M (17267584) or 18 lbs. at (+46) 17267585 and on and full
    oil of 11.3 lb. at (-14).
     
  18. Dec 15, 2018 #38

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    5,981
    Likes Received:
    1,996
    Location:
    Rocky Mountains
    Just a probably unrelated observation:

    Back in the day before fuel stops were not as common as they are today the only car I drove that I never ran out of fuel in was my 57 VW. It had no fuel gauge. A stick under the hood served to check the quantity before a longish trip. The feature that was the real savior was the reserve valve that could be operated with a simple motion of the foot when the engine started to surge. That meant I needed to get fuel soon and I knew just how much range was left. Modern autos with a low fuel idiot light now serve the same function.
     
  19. Dec 15, 2018 #39

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,028
    Location:
    Mojave, Ca
    I dont intend to get into another debacle concerning regs, but 91.205 applies to aircraft with "Standard Category" AWC. Bill's bird will be an EAB.

    Beyond that however, the more practical realization is that even a C-172 or PA-28 does not conform either. Both have indicators in the wing tanks, and both will not read accurately for some time below full. They are therefore dangerously misleading in the sense that they will indicate full with a significant portion of fuel not present. And then there is the case of the PA-22 with the optional 8 gallon aux tank... No fuel level indicator on that one, and if you didnt take the time to clear out room in the wing tank for those 8 gallons you would pump all of it overboard when attempting to transfer. Hard to believe such a system managed to make it through Type Certification!

    But before we go down the path of who can find the regs to support their position, I suggest we look at the "intent" of these regulations... And that is to provide the pilot with the means to discern his instantaneous fuel state. I think it is very shortsighted to believe that means "only" a needle that bobs up and down with the fuel level in each tank. We have many more tools available today than when my TriPacer earned its TC. And even if Bill decides to abandon part 23 or 91 entirely, there is a large body of evidence that plenty of successful aircraft do not follow the "one indicator per tank" rule.

    I fully agree that fuel system design is nothing to take lightly, but I believe we can do MUCH better than blindly following outdated concepts of fuel state INDICATIONS. There are too many examples of "compliant", but bad to borderline dangerous indicating systems on past TC'd aircraft. Bill can certainly do better.
     
    rv6ejguy and Himat like this.
  20. Dec 15, 2018 #40

    Himat

    Himat

    Himat

    Well-Known Member

    Joined:
    May 5, 2011
    Messages:
    2,752
    Likes Received:
    632
    Location:
    Norway
    Measuring flow versus tank volume have been discussed, I will add another perspective.

    Monitor pump performance.
    Measuring the current an electric pump draw should tell something about the operational environment. I would expect a low current drain if run dry and a high current drain if blocked. Between there a “normal” operating band, where it might even be possible to tell something about the fuel level in the tank pumped from by the current draw. A temperature sensor on the electric motor would tell if the pump does work as “normal”, as would a rpm sensor.
     

Share This Page

arrow_white