TECHNICAL : How to size a fuel pump

[groggory]

I thought I knew how to size a fuel pump but apparently I'm wrong. I've been having some troubles with my car and it's causing me to do lots of reading and I've fallen down the rabbit hole.

Note: I'm using my car as the protagonist in this story. This is a technical article, not an article about my car. I just wanted to put in some real world numbers so the equations can be checked against known variables.

Summary: My calculations tell me that my car, at 15 PSI, 50% Injector Duty Cycle of 830cc injectors, 3 bar FPR, and stock fuel pump should be going lean, but it's not. At that moment it's sitting around 11.75:1, which is very much NOT LEAN.

The question that I'm imposing in this thread is what is wrong with my calculations. Empirical evidence proves to me that they're wrong.

What are the right fuel sizing calculations?



Ok, here's the story...right now I'm using an OEM mk4 fuel pump

Maybe I should test out my fuel system and see.

As it stands, at 50% IDC @ 15 PSI boost on my 830cc injectors that works out to 114 liters/hour at the pump....

According to USRT I'm already overstressing the hell out of the OEM pump, which at this flow rate should max out at around 61 liters/hour!

Specs care of USRT....
	  	  	  	  	  	  
OEM Mk4 (non-R32/TT225) Intank Fuel Pump Flow Rate	
  	  	  	  
PSIG	Amps	Lbs/Hr	Gal/Hr	Ltr/Hr	Ltr/Min	CC/Min	  	  
40	 8	 145	 24.17	 91.48	 1.52	 1524.72	  	  
45	 8.4	 130	 21.67	 82.02	 1.37	 1366.99	  	  
50	 8.7	 120	 20	 75.71	 1.26	 1261.83	  	  
55	 8.9	 111	 18.5	 70.03	 1.17	 1167.2	  	  
60	 9.3	 97	 16.17	 61.2	 1.02	 1019.98	  	  
65	 9.4	 83	 13.83	 52.37	 0.87	 872.77	  	  
70	 9.7	 71	 11.83	 44.8	 0.75	 746.58	

.
..
...
....
blank line
blank line
blank line
blank line

I'm currently using Delph 830 injectors and a 3 bar (43.5 PSI) fuel pressure regulator. These injectors flow 830cc / min @ 3 bar.

According to vag com logs, when I'm boosting 15 PSI and am at around 6700 RPM my injectors are at a pulsewidth that calculates out to 50% duty cycle (pulse width / RPM * 1200)

If you scale out injector flow to a new fuel pressure (due to the boost) you can see how much each injector is flowing for me at that time. The formula is: sqrt(new fuel pressure/ old fuel pressure) * old flow = new flow. For me that is:
((43.5+15)/43.5)^.5 * 830 = 962 cc/min

Then multiply by the duty cycle to get the flow per injector so the equation for the flow per injector is (for 50% duty cycle):
((43.5+15)/43.5)^.5 * 830 * .5 = 481 cc/min

So each injector is flowing 481 cc/min @ 59 PSI fuel pressure. Ok, cool.

My old way of thinking was to size a fuel pump for a person you'd take the max flow of all 4 injectors at the max fuel pressure, add them all up, and your fuel pump needs to be able to keep up with that flow rate at that fuel pressure.

However, as you can see from the specs of the mk4 fuel pump, there is NO WAY that it's keeping up with that line of thinking. I would need to flow 4 * 481cc @ 59 PSI fuel pressure = 1924 cc/min. The OEM fuel pump can only flow around 1020 cc/min at that fuel pressure.

At the per injector viewpoint, the OEM pump still has plenty of head room left on it. But if you multiply by 4 injectors you get:

1924 cc/min @ 50% duty cycle, which is well outside the OEM fuel pump's capability. I should be going WAYYY lean at that point in time...but I'm not. In fact, at 15 PSI and these injectors the OEM pump should be unable to flow above 30% duty cycle...which means I should be going lean at anything over 30% duty cycle, which for me is anything above around 3800 RPM (where I hit around 15 PSI boost).

This leads me to believe that the way I thought you should size fuel pumps is wrong.

....
...
....


Here's my hypothesis on how to size a fuel pump...

When sizing a fuel system, how do you match up flow between the injectors and the fuel pump? My intuition tells me that you don't need the fuel pump to flow the full capacity of all 4 injectors, because there is pressure in the line and the fuel line acts as a sort of surge or buffer for the individual injector pulses. However, you can't think of the pump and injector as instantaeous devices either. So my intuition tells me you should:

Flow of injector @ duty cycle @ pressure = ______

Multiply that by a factor = ___ >100% ___

Make sure your pump can flow at the answer to that equation.

So in my case, if the factor is say, 60% then

830 injector @ 50% duty cycle @ 15 psi boost = 481 cc/min * (1.6) = 769 cc/min

So in my current scenario my fuel pump should be sized such that it can pump 769 cc/min @ 59 PSI, which is well within the capabilities of the OEM fuel pump.

Is my thinking right?

[Chickenman35]

Simple solution.

1: Pickup phone.

2: Call Scott at USRT and ask him about fuel pumps.

3. Go buy the Walpro or Bosch pump that he recommends. ( I recon at least a Walpro 225GPH or a Bosch 044 )

[groggory]

Simple solution.

1: Pickup phone.

2: Call Scott at USRT and ask him about fuel pumps.

3. Go buy the Walpro or Bosch pump that he recommends. ( I recon at least a Walpro 225GPH or a Bosch 044 )

lol.

Yes, I know what will work. Yes, I know there are people with knowledge out there. I'm trying to further my own knowledge.

I have a bosch '044 surge tank built by INA sitting in a box waiting to be installed.

The issue isn't that I don't know what I should put in my setup. I want to know how to calculate fueling requirements in general. Not just for my setup.

I'm using my setup as an example so we have real world numbers to put in the equations.

[groggory]

^^ up ^^ for a some insight.

[Shamrock]

well thats your hypothesis.. why don't you do the experiment now... I remember someone had a video of how bad the bosch 440cc injectors were.. couldnt you just setup a pump and a little fuel hose with a fuel rail and measure it? get some graduated beakers and a stop watch?

[groggory]

well thats your hypothesis.. why don't you do the experiment now... I remember someone had a video of how bad the bosch 440cc injectors were.. couldnt you just setup a pump and a little fuel hose with a fuel rail and measure it? get some graduated beakers and a stop watch?

I could... I was hoping the answer to this would already be known

Id like to know how interested people are in this answer before I got too deep in

[04 GLI Luva]

There's a reason why we have return lines

Hell I know there's gotta be at least 1 person thats prolly only making 400 yet has IE's dual 044 setup lol

[ejg3855]

There's a reason why we have return lines

Hell I know there's gotta be at least 1 person thats prolly only making 400 yet has IE's dual 044 setup lol

E85 takes alot more fuel

[groggory]

E85 takes alot more fuel

I read 30% more is a good round figure for e85

[04 GLI Luva]

If theyre on e85 i hope theyre making more then 400 lol

30% is correct +/-3%

[ejg3855]

the lack of consistency in E85 sucks, it ranges from 65% to 90%+ at times.

[Mindfault]

Wally 255, or bosch 044. And + a surge tank and ull be set. Then i believe ull be fine man, if u want a wally, ill sell u mine, i desperately need to upgrade! but this should be a good write up for people in the same situation.

[K20017]

I have the stock in-tank pump with an IE surge tank and Bosch 044. At 25psi or higher, my A/F is around 13:1. I will do a log and see what my IDC but shouldn't that support my setup?

[groggory]

I have the stock in-tank pump with an IE surge tank and Bosch 044. At 25psi or higher, my A/F is around 13:1. I will do a log and see what my IDC but shouldn't that support my setup?

I'm quite confident that you are 100% good. However, if you want to log blocks 001-031 we can always take a look. As people have already pointed out, for 90% of the BT guys on the 1.8t platform, OEM feeding surge tank + Bosch '044 will always fit the bill.

However, in this thread I'm trying to quantify things a bit more instead of just saying, that'll do it.

[K20017]

Ok I will. Anything below 25psi, I am around 11.5-12.0. My tune isn't perfect either

And for your pump size quest, I am assuming you are negating fuel line sizes?

[Atomic Ed]

This should help:

http://jayracing.com/images/products...e_A1000EFI.JPG

[groggory]

This should help:

http://jayracing.com/images/products...e_A1000EFI.JPG

Thank you. That will be one piece of information I can use. PS...you can find many more graphs like that in my Fueling FAQ

[groggory]

Ok I will. Anything below 25psi, I am around 11.5-12.0. My tune isn't perfect either

And for your pump size quest, I am assuming you are negating fuel line sizes?

Not negating fuel line sizing...That's on my radar too...but I am trying to keep the number of variables to tackle at once under control.

[groggory]

Bunp

[groggory]

Playing around with some google spreadsheets. Plotted a few flow rates for some common pumps...

Which I now see isn't shown in tapatalk

[groggory]

^^ bump

[groggory]

Does anyone have a graph of the pulses for all 4 injectors?

I have a feeling this can be solved if I knew injector overlap.

[groggory]

Does anyone have a graph of the pulses for all 4 injectors?

I have a feeling this can be solved if I knew injector overlap.

I need to learn about cams and timing advance:-). Solution coming soon

[need_a_VR6]

Two things that you havent mentioned:

- Battery voltage, the higher it is the more your pump flows. Depending on how it was flowed in te data you have provided (12v, 13v, 14v, etc) you might have some headroom

- Fuel pressure, the stock pump output drops hard with rising pressure. 3bar fpr plus 1bar boost will tax things less then even a stock turbo 1.5bar, and you have a ton more flow vs using a 4bar base pressure.

[groggory]

Two things that you havent mentioned:

- Battery voltage, the higher it is the more your pump flows. Depending on how it was flowed in te data you have provided (12v, 13v, 14v, etc) you might have some headroom

- Fuel pressure, the stock pump output drops hard with rising pressure. 3bar fpr plus 1bar boost will tax things less then even a stock turbo 1.5bar, and you have a ton more flow vs using a 4bar base pressure.

The pump data I've provided is at 13.5v

Fyi

[need_a_VR6]

Any idea on your actual voltage?

[toby lawrence]

Does anyone have a graph of the pulses for all 4 injectors?

I have a feeling this can be solved if I knew injector overlap.

I was originally going to comment on this but my brain was too foggy to wrap my mind around it.

Order a DSO Quad or something and you can measure all four signals. Based on the definition of duty cycle, I'd say your existing calculations are right... but like you, I'm also confused how they could be since you're obviously not running out of fuel as hard as your calculations would infer.

There's also the issue of open and close time on the injectors which will skew the real results compared to theoretical results. I think the simplest way to do it would be:

- get a fuel pump
- get a fuel pressure regulator
- get a gas tank and tap it and put the fuel pump in it
- set up the injector for the pressures you want it running at
- get an Arduino and set it up to control the injector at the duty cycle you care about
- get a graduated cylinder, and let 'er rip

[groggory]

Any idea on your actual voltage?

What voltage?

[groggory]

I was originally going to comment on this but my brain was too foggy to wrap my mind around it.

Order a DSO Quad or something and you can measure all four signals. Based on the definition of duty cycle, I'd say your existing calculations are right... but like you, I'm also confused how they could be since you're obviously not running out of fuel as hard as your calculations would infer.

There's also the issue of open and close time on the injectors which will skew the real results compared to theoretical results. I think the simplest way to do it would be:

- get a fuel pump
- get a fuel pressure regulator
- get a gas tank and tap it and put the fuel pump in it
- set up the injector for the pressures you want it running at
- get an Arduino and set it up to control the injector at the duty cycle you care about
- get a graduated cylinder, and let 'er rip

I was thinking of hooking my logic analyzer up to all 4 injectors and seeing what the timing looked like.

Do a wot pull, start logging the injectors at the start of the pull, stop logging at the end. Vag com blocks 1,31,115 while I'm doing that. And put a voltage shunt on the fuel pump to get the fuel pump current.

Then I'd need to use an arduino to give a fuel injector the signal and learn the open and close delays.

I think all that together may give me the answers

[need_a_VR6]

Voltage at the pump. Any idea on the hp you are at right now with your setup?

[groggory]

Voltage at the pump. Any idea on the hp you are at right now with your setup?

Nope and nope

[groggory]

I just had a huge DUH! moment.

I've been thinking that with increased boost you get increased injector flow due to your fuel pressure regulator being 1:1

TOTALLY WRONG!!!

I forgot that your FPR is a 1:1 rising rate regulator because the intake manifold is pushing up as hard as the gas is pushing down...due to the boost.

So if you have a 3 bar FPR and are are boosting 2 bar you have...
* 5 bar fuel pressure
* 2 bar 'boost' (intake manifold pressure)
* The fuel injector will continue flowing at 3 bar rate, as you would in open air. This is due to the idea of differential pressure.

WOW!! I overlooked a huge concept.

:-) More thinking to come and perhaps some conclusions after I put my thinking cap on some more

[groggory]

Ok, this may help some...

http://forums.vwvortex.com/showthread.php?1160233

and

http://www.stangnet.com/mustang-foru...rocess.736061/