How To - Understand Fuel Tanks
 

How To - Understand Fuel Tanks (Last edited 21 August 2014)

This is another bit of work in progress and there are probably things I've got wrong. Hopefully these will be corrected through informed comment. I'll edit it as additional info is gathered and mistakes are corrected.

Over the years, I’ve read loads about what happens in the fuel tank if for example, a diesel pump fails, or Fuel Filter Syndrome (FFS) develops in a petrol system. Thankfully there are only a few different problems, but they do happen frequently and they get discussed a lot. However, I’ve often felt the explanation of cause and effect is little more than guesswork or misinterpretation. So, I decided to learn what I could about the way tanks work and did some research. I admit I was surprised. Boiling it down, I thought a simple fluid flow diagram for the fuel tanks would be useful. Basically, if you understand the flow paths you can explain the things that happen. What surprised me is the fact that all fuel tanks contain not one, but three pumps.

In addition to the electric pump located in the RHS of the tank, there are two static pumps located deep inside the plumbing system. These transfer fuel within the tank and ensure the electric pump inlet is always submerged, no matter what the fuel inventory or the driving conditions. Now this seems contrary to many things I’ve read on the forums, especially regarding diesels. The ‘quarter tank’ rule for diesels has always intrigued me and I now see the explanation of it. Similarly, the idea that under certain circumstances the RHS of the tank will have all its contents transferred into the LHS seems an illogical design feature. So I made some drawings of the petrol and diesel tanks to simplify the workings.

https://www.mediafire.com/convkey/fc...zg2t87m56g.jpg


https://www.mediafire.com/convkey/30...n0o49l4a6g.jpg
These show only the liquid flows. Vapour and pressure handling are omitted. The saddle tank and interior pipework are shown diagrammatically, so all this stuff is a bit different in reality. The idea behind the internal plumbing becomes apparent when the fuel level is below the central dividing hump. It makes most sense when the fuel inventory is very low, so the drawings show that situation. Diesel or petrol, the system is identical in principle, but the way it’s done is slightly different.

The primary electric pump, whether diesel or petrol, is a centrifugal (impellor)type. The diesel one generates around 30psi at the outlet when in top condition. The single petrol pump generates around 50psi. The UBP in the diesel is a positive displacement type with a 'pigtail' gear design.

A centrifugal pump can run against a closed outlet while maintaining its delivery pressure, but it must be adequately primed. To ensure this, the pump is fed from a permanently flooded pickup or swirl cup. It’s the job of those other two pumps to keep the swirl cup full. When there's plenty of fuel in the tank, the electric pump and its swirl cup are completely submerged, so while there are transfers happening within the tank, they only become important when the inventory falls. At very low inventory, any air getting into the pump inlet is vented off at the upper end of the pump through a scavenger/priming valve which closes when the pump is hydraulically locked. Make note of this valve.

The two secondary in-tank pumps are static inductors based on the Venturi principle. These get their motive power from pressurised fuel generated by the electric pump (petrol) or the diesel’s HP injector pump back-flow from the engine. Their sole purpose is to ensure the swirl cup remains full under all conditions. Since they will only work when pressurised fuel is fed into them, any circumstance which reduces this pressure will cause a tank malfunction. For example, the diesel HP backflow passes through the cooler mounted externally near the tank. A fuel leak from the cooler may inhibit delivery from the inductors. Similarly, FFS in the petrol tank may inhibit feed to the inductors. This will also cause tank malfunction and explains why FFS is sometimes accompanied by an absence of fuel in the RHS.

Diesels with two electric pumps (ITP and UBP) may under certain conditions fail to get sufficient fuel if one fails. The ITP has adequate delivery capacity when assisted by the UBP. So, when it has to work alone, it needs to do the work of the UBP as well. Whether it copes with the load depends very much on the lift it has to cope with. The lift of a pump is determined by the outlet pressure. The required lift is the vertical height between the fuel surface in the tank and the HP pump inlet on the engine. While the height remains below some critical point, fuel will get to the engine. The only variable in this equation is the fuel level. As the level drops, the required lift increases, until a point is reached where delivery fails. I assume this is at the nominal quarter tank level. As fuel delivery reduces, the volume of HP backflow from the engine will fall. This backflow is vital for the correct handling of fuel by the Venturi pumps. As noted, these scavenge fuel from the left side of the tank into the right and from there into the ITP swirl cup. Clearly, if the scavenger pumps are inoperable due to lack of HP backflow, the RH side of the tank will run dry.

If the ITP fails before the UBP, this has to lift fuel from the tank and deliver it to the HP pump on the engine. I would guess then, that the UBP is a different type from the ITP and is logically positive displacement. (This has subsequently been confirmed and there are some nice photos (somewhere) of the interlocking spiral worm gears). It must have sufficient 'suction' to raise fuel from the tank (via the dead ITP) on startup. It would be interesting therefore to compare the innards of each. Again, the quarter tank rule seems to apply, so that the UBP can cope with the lift and engine requirements. As the UBP eventually fails, the scavenger pumps fail and fuel seemingly becomes transferred the LHS of the tank.

Later diesels only have a single primary pump (ITP). I believe this is a higher spec'd type and presumably has the lift and capacity to cope with the demands of the engine and the tank scavenging system. When this pump starts to fail, you'd expect the outcome in terms of fuel distribution to follow a similar pattern.

It would seem logical in cases of double pump failure to replace the ITP with the later type and simply bypass the UBP. Apparently this isn't possible, I assume because of design differences between the two. (??)

The following discussion raises the question of the fuel cooler fitted in the diesel system. This isn't shown in the above drawing to keep it simple, but there are some implications if it goes wrong. An example of this can be found here: https://www.the75andztclub.co.uk/for...d.php?t=193016

TC

Excellent post T-C, you say it's a work in progress and subject to revision. The How To section is deliberately locked to keep all the threads uncluttered and clear for reference purposes.

When you are happy for it to be placed in there, let a mod know, and we will make a copy of it to be kept there for future reference. Your live thread will of course remain here for comments and amendments.

Good start TC but I think you need to add the fuel cooler, thermostatic valve and the associated pipework to the diesel circuit, because when the fuel is cold I don't think it returns to the tank but stays in a closed loop circuit, with the ITP topping up what is taken by the HP pump to fuel the engine. Overspill from the ITP returns to the tank during this time.

If you look closely at the circuit there isn't a simple bypass of the fuel cooler to the tank, and the way it is piped, taking into account the different pressures from the returning fuel and the ITP, it diverts fuel straight back to the UBP which gets it up to temperature in cold climates alot faster, then as everything warms up the thermostatic valve starts to open to regulate fuel through the fuel cooler and back to the tank.

One other point is that with a failed UBP the engine stops. The ITP can't supply fuel past the UBP. A good UBP will run fine with a failed ITP with enough fuel in the tank, although it has more work to do.

The system is a bit more clever than it looks!

Mike

Indeed. Thanks for your comments. I've tried to give rudimentary insight of the relationship between tank contents and the pumps in so far as we are affected by their failure (through wear or FFS). The basic flow diagrams explain the principles of fuel flow to and from the primary pump and how the fuel is distributed by the venturis. I fear it will be indigestible anyway, so I'm lothe to go into greater detail. The title of the piece may well be over ambitious, I'm certainly no tank expert and there are things about it I left unread. It just seems to me that tank/pump issues are discussed a great deal without any real insight on what's actually happening in the tank. Many threads on this topic assert that fuel is pumped from RHS to LHS because that is what seems apparent. The 'quarter tank' rule is another example. What I hope is that what we see happening during pump failure and the anecdotal fixes for it have a logical explanation. Whether this does that remains to be seen.

EDIT: I did think twice about the results of ITP and UBP failure on fuel delivery, but I feel certain I've read of engines running on the UBP alone. I'll have to go back to that for sure.

TC

Yes my UBP did 20k miles on its own that I was aware of, and probably a fair few miles more that I didn't know about.

It was your quote here that maybe I misunderstood:

[Quote] Diesels with two electric pumps (ITP and UBP) may under certain conditions fail to get sufficient fuel if one fails. The ITP has adequate delivery capacity when assisted by the UBP. So, when it has to work alone, it needs to do the work of the UBP as well. Whether it copes with the load depends very much on the lift it has to cope with. [QUOTE]

Mike

Quote ' I would guess then, that the UBP is a different type from the ITP and is logically positive displacement. It must have sufficient 'suction' to raise fuel from the tank (via the dead ITP) .'

Open to correction here, but surely the UBP is the same as the ITP, Both look the same type, long cylindrical. The UBP is of the double interlocking helical type. That cant be positive displacement ? or can it ?.

They are different designs - as you say the UBP is a type of screw pump but the ITP has a rotary impeller.

The ITP is self priming due to a normally open ball valve situated at the top of the pump, allowing diesel to flow in. This closes as the pump pressure builds up.

This means that with a failed ITP the ball valve is open so the UBP doesn't actually have to suck fuel through the ITP itself. It also means that when the fuel level drops down below the top of the ITP the UBP will start sucking air. Another clue to why the car can stop with a quarter tank of fuel onboard after the ITP has packed up.



Mike

Thanks Mike and Derek for your input. As more insights are gathered, I'll amend the account to make it more accurate. However, I don't want to make it more complex. The important thing is that the basic mechanisms are explained without too much detail of how they're managed. I see that the IPT/UBP inter-relationship does need some work.

TC

Very well done T-cut, it is the first time the fule system mystery has made any real sense to me.

A couple of questions....

I don't think the ITP pump type would generate anything like 52PSI, so why a pressure regulator there and are you suggesting that the regulator is actually in the tank? If it is indeed there, between ITP and UBP - why there, surely it would have been more sensible to have a regulator after the UBP?

The T, where the HP return flow goes back to the tank - I don't a simple T piece would work to share the flow between the two venturi pumps would it, without some extra help to force it to share the flow?

When that ITP does fail, might a second UBP in the line from the tank, make up for the failed ITP? The UBP is perfectly capable of self priming I found, from bench testing one.

OK four questions :D

Must admit I had not thought of venturi pumps to pump fuel from one side of the tank to the other. So I suppose due to the venturi pumps needing pressure to work, that the return system and cooler must be under some pressure?

You're right. The combined pump delivery pressure to the engine compartment is set to 36psi. The pressure regulator mentioned sits inside the so-called 'filter adapter' module located in the LHS tank port. Here's Rimmer's drawing (linked direct from the site)

http://www.rimmerbros.co.uk/ItemImag...GRID001964.gif

This is Item 5. The pressure regulator is calibrated to 52psi, but remains shut under normal operation. Exactly what abnormal event would open that I've not figured out. I think Mike will know. I've amended the text to reflect the correct spressure.

As above, it's function isn't yet clear.

The drawing is simplistic and as noted doesn't reflect the exact plumbing in the tank. The basic principle is all it illustrates.

I guess it would probably work. Maybe somebody will try it.

The return line from the HP/Injector system via the cooler and into the scavenger system is under pressure otherwise it wouldn't work. I'm guessing that if volumetric flow to the engine is down (due to failing ITP) then the scavenging system will also fail. This explains the observed fuel levels in the tank when this happens.

TC