Upgrading fuel filtration on an IDI VW

[TurboJ]

AGAIN, for those of you who missed it in the first EPISODE:

you can feed the injection pump with both VACUUM and PRESSURE..

pressure does not change the timing advance curve..

That was not my question, but thank you again.

[TurboJ]

Run a separate pump into a fuel cooler rad, that all its job is to cool the fuel and return it to the tank. Then run your filter, water separator, inline pump, final filter to the IP. Send the return back to the tank. Two separate fuel loops, but you will be able to run the fuel on a temp control and keep it exactly where you want it. The cooling loop would be able to transfer more turnaround and should be more effective.

This is not a possibility for me unfortunately.
Two separate loops would necessitate a four-connector sender unit. If the feed line and return line coming from the injection pump would begin from the in-tank sender unit, then so would the cooling loop's lines. Because if they did not, then the hot fuel and the cool fuel wouldn't mix very much.

As I have a sender unit made already and it has two lines on it, I have to incorporate the cooling circuit into the main system.

I will have to use some sort of a filter on the VE pump's return line, though.
Thanks for reminding about the contaminants, TheMan53! I would think putting it into the pump's return line
instead of the feed line so there would be no pressure loss on the pump's inlet.

Anyone know what kind of pressure there is on the VE pump's return line just after the pump?

The idea of the catch tank in my design is to

1. Eliminate the possibility of air entering the VE pump
2. Minimize the load on the VE pump as it sucks fuel in
3. Minimize the load on the transfer pump as it sends fuel to the VE pump
4. Allow the right amount of fuel feed to the VE pump at all times

...the #4 is important - knowing exactly the required amount of fuel flow for my heavily modified pump is very difficult, and could change under different circumstances.
A transfer pump that would flow enough (but not too much) would be necessary without the catch tank.
A big pump, at that. On the catch tank design, the transfer pump only needs to pump as much fuel as the engine is burning at any given moment.
Without the catch tank, the transfer pump needs to pump as much as is flowing through the injection pump at any given moment.
Also, since the transfer pump would need to flow more fuel than needed by the pump (to make sure it's always enough), it would pressurize the entire fuel system and that would increase heat in the fuel.

Let's say I add a rough filter to the return line from the VE pump, after the fuel cooler.

Can you please tell how you see the system working as a whole?
How much fuel would be flowing back to the fuel tank from the catch tank?
How much of the VE pump's return flow would go to the fuel tank, how much straight back to the pump itself ?

[TurboJ]

So this is my latest design evolution:





Now all the fuel should be filtered.

What do you think, how will the fuel heat spread around the system on this design?

[Alcaid]

Why the rough filter I.? Afraid of debris from the cooler to get into the pump by taking the shortcut through the surge tank?

I can recommend these filters as rough filter, see-through so easy to see if there is any debris in it, cheap enough to have several on the shelf
http://www.ebay.co.uk/itm/5-x-LARGE-IN-LINE-DIESEL-BIO-VEG-FUEL-FILTERS-/310270143508

[TurboJ]

Why the rough filter I.? Afraid of debris from the cooler to get into the pump by taking the shortcut through the surge tank?

That, and debris that might be generated inside the injection pump and/or the injectors. However unlikely..

Thanks for the link, BTW!

[Alcaid]

I had a look at a MTU 2000 series engine today (20 litre, 900kW, boat engine) It used 2 ea. 002 092 19 01 fuel filters in parallell, so 450kW each

[theman53]

Why couldn't you just plumb into the return line from the pump? I would ditch the catch tank. Right before fuel enters into the tank on the return, plumb you cooler in and then dump it into the tank. It will not pick up any heat from the line to the pump, so as long as your cooler is efficient enough to cool it should work...if it would work up front it should work in the back right? Also I don't get why you have the catch tank after the filter. I would think ideally you should have the filter as close to the IP as possible, especially if you are sending the return back through it. These filters are not 100%, so you are basically saying if it makes it past the filter the first time it should be ok to keep it going though until it gets sprayed out the injectors. I am all for overkill, but if this is needed doesn't AKI have a formula already laid out and you could copy? Does he have a range that the fuel temp should be in? That is what I would be more interested in and then see if I could hit that number stock or what is needed at that point to do it.

[TurboJ]

Hmm. Everyone is so eager to tell me not to use a catch tank.
Yet I see it as a huge improvement...

The catch tank is after the filter because the whole point of the catch tank in this design is to eliminate the flow loss and pressure loss of the filters so
the injection pump can work as efficiently as possible. Normally it would have to suck all the fuel through a filter and that is a restriction in every case.
In this design there is the possibility to use a more efficient main filter as well, since that will not limit the flow to the injection pump like it would
without the catch tank.

The catch tank design means practically no resistance in the inlet line of the injection pump, AND at the same time it completely eliminates the possibility
of air entering the injection pump.

See, if the catch tank is not pressurized, yet it's always full of fuel, there is practically zero vacuum on the injection pump's inlet line at full flow.
That means there's practically no resistance, and not even a possibility of leaking hose connections letting air in.

Not straining the VE's feed pump and completely eliminating air, are, I think, very important improvements.

Now, since there's pressure on the injection pump's return line, this will overcome the drag on the cooling radiator, and that of the secondary rough filter (which
catches the particles that might be generated inside the pump and injectors.

I get the point of only a part of the fuel going back to the fuel tank and passing the main filter again on its second run.
But the ratio for how much fuel is sent back to the fuel tank and to the second pass through the main filter, is governed
by the capacity of the transfer pump, and engine load. The more the transfer pump flows, the more fuel will be sent back to the fuel tank
from the catch tank. The more engine load, the less fuel will be sent back to the fuel tank
from the catch tank. So the flow rate of the transfer pump is what really counts here.

I see it like this.

Assume the injection pump flows 240 lph. At max load it sends 60 lph of that to the injectors. Hence, anything from 180 to 240 lph is what is returned to the catch tank.
Assume the transfer pump flows 120 lph. At max load, then, there is a total of 300 lph flowing into the catch tank. Only 240 lph can go to the injection pump. That leaves 60 lph
going back to the fuel tank, which increases to 120 lph at idle. So, at any given time, 1/4 to half of all the fuel is going back to the fuel tank and through a rough filter and a main filter.
In addition to this, ALL fuel is going through the secondary rough filter on every cycle.

If a 180 lph transfer pump is used instead, however, the amount of fuel that flows though the main filter on every cycle, grows from 25-to-50 percent to 50-to-75 percent.
A 240 lph transfer pump would yield a rate of 75-to-100 % on each cycle.

An original VAG fuel filter is a 15 micron unit. If I use a 5 micron unit instead (which I can because of the non pressurized catch tank), every cycle will filter a lot higher proportion of the contaminants in the first place. Assume twice the filtration, but half of the cycles though the filter. With the effect of the secondary rough filter, I doubt filtration would be any kind of a problem.

Now, one interesting alternative could be fitting the main filter between the catch tank and the IP, and then fitting a pressure regulator to the catch tank -to-fuel tank overflow line.
Then, adjust the pressure regulator to accurately negate the vacuum caused by the main filter. The only problem with this design, of course, would be that you'd have to re-adjust the regulator as the filter becomes more and more clogged by particles.

[TurboJ]

EDIT:

Thank you theman53 for pointing out the filtering flow problem. I didn't really think about that as much before this post.
I think I will fit a very high flow transfer pump. This way I will get MORE filtration runs than there are runs from the IP back to the catch tank.
That together with the 5 micron main filter should dramatically improve filtration at the same time as the flow is increased.
And still with zero pressure loss on the IP's intake.

Too bad this way the otherwise possible longer filter change intervals have to be forgotten.
So I'll have to use some BIG filter unit in there.

[R.O.R-2.0]

EDIT:

Thank you theman53 for pointing out the filtering flow problem. I didn't really think about that as much before this post.
I think I will fit a very high flow transfer pump. This way I will get MORE filtration runs than there are runs from the IP back to the catch tank.
That together with the 5 micron main filter should dramatically improve filtration at the same time as the flow is increased.
And still with zero pressure loss on the IP's intake.

Too bad this way the otherwise possible longer filter change intervals have to be forgotten.
So I'll have to use some BIG filter unit in there.

why are you afraid to pressure feed your pump?

the dodge guys feed there bosch VE pumps with like 30psi, and lots of GPM...

and run huge fuel lines..

they run 100% of the fuel thru the pump..

there is no catch tank, and some of those guys build near 1000hp...

so, tell me that your catch tank is necessary? what is its purpose? why cant you feed your pump with pressure?

VWs are about the only diesel that dont pressure feed their injection pump STOCK..

im just trying to figure out why you cant pressure feed your pump as well?

Giles and Tyler fed a Bosch VE w/ 150psi once.. didnt change anything, but it was very easy for the pump to pull fuel..

[TurboJ]

I have no reason to believe I should pressure feed the pump.

As to the reason for running a catch tank, I have explained this about four different times on this very thread.
Just because some people run a certain type of fuel system doesn't mean it couldn't be improved.

If a pressure feed to the pump was necessary (which it is not),
that could be accomplished with the use of a pressure
regulator in the catch tank's overflow line.

[R.O.R-2.0]

I have no reason to believe I should pressure feed the pump.

As to the reason for running a catch tank, I have explained this about four different times on this very thread.
Just because some people run a certain type of fuel system doesn't mean it couldn't be improved.

If a pressure feed to the pump was necessary (which it is not),
that could be accomplished with the use of a pressure
regulator in the catch tank's overflow line.

i just think you are making things overly complicated...

[R.O.R-2.0]

pressure feeding the pump doesnt cause any of these:

1. Eliminate the possibility of air entering the VE pump
2. Minimize the load on the VE pump as it sucks fuel in
3. Minimize the load on the transfer pump as it sends fuel to the VE pump
4. Allow the right amount of fuel feed to the VE pump at all times

i dont understand why you want the pump to be under vacuum?

even tho it will be pulling fuel a short distance, why would you NOT want to pressure feed it?

it will still be PULLING fuel, even tho the catch tank is close.. but there will still be vacuum, unless you plan to mount the catch tank ABOVE the pump some how?

the possibility for air to be in a pressurized liquid is less than one under vacuum..

if you want to minimize the load on the pump, then pressure feed it..

if you want the right amount of fuel on hand at all times, again, pressure feed it.

what do you have against pressure feeding the pump? its been proven millions of times that it works great for high hp Bosch pumps..

[TurboJ]

1. Eliminate the possibility of air entering the VE pump                            -->  Requires a catch tank (the only possible 100% solution )
2. Minimize the load on the VE pump as it sucks fuel in                            -->  Pressure feed alone would suffice
3. Minimize the load on the transfer pump as it sends fuel to the VE pump  -->  Requires a catch tank  ( a non-pressurized one at that )
4. Allow the right amount of fuel feed to the VE pump at all times             -- > Pressure feed alone would suffice  ( IF the transfer pump has an internal pressure regulator and a bypass valve, AND operates at just the right pressure range )


...So you see of the four above requirements, a pressure feed without a catch tank would only meet two, at best.

As for the location of the catch tank, obviously it will be fitted higher up than the injection pump.

If a pressure feed is a necessity on these engines, why do most VE-pump equipped cars in the world run without?

----

I don't much care and for continuing to arque about how bad a catch tank setup is even though the majority of race cars use them.

However, I would like anyone's help in this question:

How much pressure and how much force is there in the VE pump's return line?
I would like to know how much resistance that flow can overcome.
I need to fit the fuel radiator in that line, and also possibly a secondary rough filter.
So please help me on this question!

[theman53]

as far as I know there is 0 psi in the return. The fuel hose isn't ever really full on my Giles pump. You would have to block it to get pressure. It flows a ton, but I do not think it is under any pressure

Like I asked before, what does AKI already do?