PA 28 Selector Valve

[My names Turkish]

I love flying the PA28(Cherokee) but am can never fully relax as I worry about forgetting to switch tanks, which if you check the NTSB website for example www.ntsb.gov , You will find is a very common accident for this type.

Which leads me to my question. Having flown a PA34(Seneca 1) for the first time recently, I was very suprised that it did not have switchable tanks. I wondered why then, is there a *potentially* very error prone switchable valve on the Smaller Piper. Other light aircraft I have flown, like the Cessna 152 and 172 have a valve that slects from both tanks except in an emergency situation you can manually select a tank.

Having had limited studies outside of flying in Human Behaviour and JIT, I was always told that if it was possible for a human to make an error they would and that the aim in manufacturing was to make it error proof, and I am suprised this principal has not been carried over to all aspects of airplane desgin. Or am I missing something?

[Genghis the Engineer]

I must admit that whenever I fly a PA28, particularly one with a 4-tank system, I ask myself why the heck this couldn't be automated in some way - since you never want to give the engine opportunity to "cough", you inevitably leave perfectly useable fuel in each tank.

There is at-least one good reason not to draw from multiple tanks - if you have an imbalance and empty one tank then the engine will stop despite fuel remaining in the other due to the system naturally trying to suck air from the empty tank. However, if you have tanks on the same level, and geometry such that you consistently feed from the lowest point of a combined (and linked) tank system then you can

Cessna 150/152/172 fuel systems have a balance hose (actually two, the lower feeding the fuel pump from it's mid point), so effectively only a single tank - this is particularly easy to do with a high wing system so there's a gravity-feed element to it which also effectively makes the length of hose between the tanks and the engine a form of crude "header tank". A few types genuinely do have header tanks - albeit mostly high-wing homebuilts.


So what's the solution for a low wing aeroplane like a Piper? In my opinion, link the tanks at their lowest points, and feed from the low point of that link hose (ideally via some kind of sump with a couple of litres capacity below both tanks) into the main engine fuel pump. It would be hard to retrofit onto a PA28, but very easy to design-in to a new low wing design.

G

[Keygrip]

Genghis - a "four tank" PA28?

I have never, ever, heard of a four tank PA28 (but that, by no means, dictates that there isn't one).

I used to fly a four tank (indeed, six tank) PA31 (commercially) - but have never heard of a four tank PA28.

Tell me more.

I was in the same Seneca as "Turkish" and, have to admit, it's a good question (as our conversation ensued).....WHY??

At least with the PA38 they moved it into the instructors eye line - but why a selectable tank?

Having said that.....a (rather gorgeous) aquaintance of mine did run short of fuel in a Cessna 172 at "half endurance" - when fuel only came out of one high wing tank and not both.

It may be easier, at 4,500 hours, to detect a roll due to weight imbalance than it is at a brand new PPL 45 hours.

[My names Turkish]

Gasil newsletter:

http://www.caa.co.uk/docs/33/srg_gad_gasil4of2003.pdf

In Particular Pages 22 and 30. I wish I could find my notes from Uni for some quotes, but again the things I was taught about Human Behaviour and interfacing with machinery, and designing to "mistake proof" seems to be in very sharp contrast to what we see in those pictures.

I do hear what you are saying though about the PA28, its 25+ years old so the technology is that of 25+ years ago and a similar situation would probably not occur on a newer Aircraft?

[Genghis the Engineer]

I've only flown one 4-tank PA28, it was a PA28-235c (Cherokee wing, 4 tanks, fixed gear, variable pitch prop, 235 hp Lycoming). Cruised about 140kn, "sipping" fuel at about 40 litres/hr (that's about 11 US gal/hr) when leaned for cruise !

The fuel system was similar to the conventional 2-tank system, except that there were 4 small fuel gauges partially obscured by the yoke, not 2. And of-course, the (not in clear sight) tank selector went 5 ways rather than 3. The risk of selecting the wrong tank, particularly if cruising low-level and thus trying to keep your eyes out of the cockpit as much as possible, was in my judgement quite high. (Well I did it once anyhow!).

I agree that the PA38 fuel management panel is a huge improvement on anything fitted to any PA28 I've seen from the viewpoints of visibility, clarity and ease of use, or for that matter most Cessnas. I know it's high and obtrusive, but why it couldn't be incorporated into later PA28s is beyond me.

G

[despegue]

It is good airmanship to switch tanks each 30 min on the PA28, and WRITE DOWN the time of switch and to which side.
Fuel inbalance should be detected during your 15min FREDAS checks.
The C172 is indeed quite prone to an imbalance due to restricted suction. If you have a situation where one tank is emptying and the other one not, a probable cause is the wrong fuel-cap on the wrong wing. As the Cessna fuel-system is "gravity-feed" based and doesn't posses a pump, the port wing has an excess-fuel drain/pipe. The starboard fuel-cap has a small air-hole in it. This way, no tank wil be vacuüm. change-over the fuel-caps and one will suck vacuüm and therefore not feed.

enjoy the end of the year...

Despegue

[My names Turkish]

Most people do write down the times of changes, but a common theme in the accidents mentioned above is a series of distractions e.g Bad weather, Other Mechanical Problems etc etc. High workloads seem to be a factor in some of these, but I suppose thats a whole other problem...

[Tinstaafl]

The Cherokee 6 has a 4-tank system. Bonanza has 2 tanks but can be fitted with tips. Aztec has 4 tanks but all of them are 'mains' & are useable in all flight conditions, unlike many other multiple tanks systems.

C310/402 has one of the more complicated systems: Each wing has...

'main': must be used for take-off & landing etc

'auxillary' : of which there were two size options that could have been fitted, only useable in level flight AND the one on the engine failure side becomes unuseable ie dead weight - it can't be x-fed. Must use 30mins (or 60mins if large aux's) from the mains before selecting auxilliaries.

Optional 'slipper tanks' behind the firewall(s). Can be none, one or two fitted to the aircraft. Must be pumped into the same side main to be used. At least they remain useable for x-feeding. As long as the transfer pumps work...

It meant that for max range flying you would have to check that the slipper X-fer pumps worked as soon as space was available in the mains (watch for a quantity rise in the relevent tank) then burn from the aux's when space became available, then back to mains, burn an hour then switch on the slipper's transfer pump(s). Aux's burnt before slippers because they become dead weight if the same side donk stops.


Aerostar was simpler but still had its traps: A belly tank that both engines fed from + a wing tank for each engine. Normally each engine fed from both the belly & its wing. Only the belly was guaranteed to maintain a fuel supply so it was essential to ensure it wouldn't run dry. X-feed would cause the engine to only feed from the opposite wing so a double x-feed could be used (in cruise) to preserve belly tank fuel.

[planecrazi]

About 14 years ago now, I used to fly a Navajo and a Cessna 421 B and seem to remember the Navajo being a rather complicated fuel system, specially on the crossfeed. I am open to correction now, as I fly the Airbus A340 for the last 4 years, but the Navajo was "dead engine, dead tank" and therefore the selector for the dead engine went to off and the live engine was to crossfeed.

The Piper range seemed a little more involved than other aircraft. One always had to think more than twice before moving the selector valve in a Piper.

The C310, long nose was another story when swopping between mains and Aux and I seem to remember that the tip tank was the main which had to be selected for take-off and landing. Sometimes I would forget to change from the aux back to the main whilst cruising in the Namibian sun and I would get the odd hiccup of an engine coughing from a dry tank, which was a big wake-up call.

As we couldn't change the fuel systems and had to live with them, it gives one and open mind for all the other aircraft you will come accross in your career, so there is a positive point in all of this. The experience learning curve just keeps bending.:D

[vector4fun]

Having spent several years flying a PA-32, I found that a cheap ($10) kitchen or oven timer works wonderful for managing fuel on long flights. Wind it up to an hour, when it goes "DING", switch tanks and wind it for another hour.......;)

[Steamhead]

On PA28
You can do your own check with a stop watch, but time from running out of fuel on one tank and switching to another and the engine resuming normal output is 4.5 seconds.
To this you must add the pilots response time to it going rather quiet in the cockpit.

Regards

[Keygrip]

Now it makes me wonder why the fuel feed on a (say) Cessna 172 IS selectable - but not on a 150/152 (apart from 'off').

[Tinstaafl]

Sometimes I like being able to both select individual tanks or both. Both, because it removes an item of work, and individually, because I can run one wing heavy (within a/c limits) if an aileron tab isn't quite right.

[Keygrip]

Tin - not questioning the benefits of being able to do something - just curious why some deliberately aren't available.

What are the pro's and con's of both systems?

[Genghis the Engineer]

Most of the current aircraft certification standards prohibit a fuel pump drawing from more than one tank at a time. So far as I know this is for the reason I stated earlier - if an imbalance (say) leads to one empty tank and one that is partly full, the pump will tend to draw air from the empty tank causing the engine to stop with fuel remaining on board.

This wasn't historically the case, and since aircraft are usually allowed to continue operating subject to their original safety requirements older aircraft are able to continue flying legally with a "both" option on the tank selector valve.

I don't think now I'd be able to certify a new type anywhere in Europe (at-least) with a both selection available to the pilot.

G

[Tinstaafl]

Gravity feed has a very good pro: Simplicity. ON or OFF. No problem with misfeeding.

Con: Can't balance fuel, can't run different fuels in different tanks (Would Sir prefer Avgas or Mogas?)

Like Ghengis says, it's not so simple with pump supplied sources since the pump suction will thry to suck the least resistance hence single source selectors. Can't see why a balancing port between the tanks wouldn't fix that though.

Maybe they expect everyone to fly continually slipped so one tank drains into the other which then overflows?


Some high wing types had the best of both: OFF, L, R & BOTH.

[Splat]

The one I fly, you may as well bin the guages as they are worse than useless. I'd like to know how anybody can detect a fuel imbalance in flight on type? Maybe the later ones are better. The only way to manage fuel is by stop watch from a known starting position.

Splat

[Keygrip]

Splat - it's when the aircraft starts to roll (and it didn't do before).

[Tinstaafl]

Provided the a/c's lateral trim is adjusted correctly to start with then an imbalance can be detectable. Has to be a fair imbalance though. The smoother the conditions the easier it is.

The flip side is that an a/c with a slightly misadjusted tab can have the roll effects reduced by deliberately running asymmetric fuel quantities.