Fuel Pressure Drop When Hot

[z-ya]

Here is a problem I am seeing on the track car after a long session. It is fuel injected NA L28, and I am using the stock fuel rail and FPR.

 

When I start the track session, the fuel pressure is stable around 38psi at idle. Near the end of the session, I start hearing a little detonation, so I have to ease up a bit on it. When I get back to the garage, the fuel pressure is 30psi at idle.

 

I would think that as an engine heats up, it will pull less vacuum, but this would cause the fuel pressure to rise when it got hot.

 

The stock FPR is a few years old (one of the nicer looking items under the hood), and I have tried two fuel pumps just in case the pump was getting hot. I don't think both pumps are having the same problem.

 

I am running a two pump system, one to keep the surge tank full (with a return back to the main tank), and another to feed the engine (with a return back to the surge tank).

 

I plan on relocating the FPR to the passenger firewall to see if that helps.

 

Comments appreciated.

 

Pete

[Tony D]

What is the temperature of the fuel in the tank?

With todays' fuels and the pressures they operate at, there is a possibility of inlet/suction side cavitation of the pump from gas bubbles flashing in the inlet to the pump---and that drops fuel pressure.

 

Is the surge tank pressurized to 2-3psi, or is it basically a 'atmospheric' NPSH that is going to your main pump? May want to restrict the bleed-back line from teh surge tank to the fuel cel to elevate the pressure there and see if that helps.

 

If it does, then it's a fuel-heat/cavitation issue.

[z-ya]

Hey Tony,

 

I do notice that the surge tank gets warmer as the day goes on. It is warm to the touch at the end of the day. The problem also appears to happen later in the day also.

 

The pressure in the surge tank is close to atmospheric. The surge tank ports are as follows:

 

From main tank - 5/16"

To main tank - 5/16"

To high pressure pump inlet - 3/8"

From fuel rail - 5/16"

 

Maybe I should reduce the return port to the main tank to 1/4"? This should allow some pressure to build up in the surge tank.

 

Thanks,

 

Pete

[cygnusx1]

Long shot, but I have seen this happen before with incorrect gauges. Is your gauge reading incorrectly when hot? I have seen sealed gauges not rated for high temp environments in engine bays. If the gauge is sealed, when it heats up, it will read low due to internal pressure in the housing. Especially a problem with liquid filled gauges in engine bays.

 

"For liquid-filled pressure gauges that have uncompensated cases, accuracy errors can be introduced even with as little as 10 degrees F change in temperature. Under this condition, the potential for error is a result of internal case pressure build-up from an increase in temperature; this will cause a downscale pointer shift (the opposite will be true with a decrease in temperature). This potential for case pressure error will be present in pressure ranges of 100 psi and lower. When harsh (corrosive or dusty) conditions exist, hermetically sealed and/or liquid-filled pressure gauges should be specified. Either of these designs will prevent foreign elements from entering the case and adversely affecting gauge operation."

 

Source: Experience and http://www.impomag.com/scripts/ShowPR.asp?RID=4779&CommonCount=0

[z-ya]

Dave,

 

You might be right, as it is a liquid filled gauge, and it is in the engine compartment. But even after sitting with the hood off for 1/2 hour or so, the pressure still read 30psi. But I am hearing detonation under full load when this happens, so I do believe that the fuel pressure is dropping.

 

Thanks,

 

Pete

[Tony D]

What you may want to consider is to install some sort of metering valve in the return line off the surge tank, that way you could test various pressures.

 

Come to think of it, simply putting a backpressure regulator for a HOLLEY on that line and adjusting it to keep 3.5psi in the surge tank would be pretty much set and forget...if you happen to have access to one of those.

 

I don't know how the Stillen Z32 regulated pressure in their surge tanks, but I know they had four pumps for each bank of cylinders! I'm thinking by looking at their tank, they ran 'some' pressure, and quite possibly more than one would suspect to help push any entrained air up and out of the top. Pressure imparts flow through an orifice...

 

I had a Carter N/A V8 pump for my surge tank, it was set up to automatically regulate to 3-4psi, and I could hear it shutting off at Idle (high bypass back to the surge tank, probably overwhelming the 1/4 line I had on there). At WOT I'm sure I never shut off, and have no idea what pressure I was running in the tank then, but at idle mine would climb to 4psi at least.

[z-ya]

I think John (gnosez) has a spare low pressure regulator. We'll give it a try at the next event.

 

Thanks,

 

Pete

[Daeron]

What about simply routing the fuel rail return line back to the gas tank rather than the surge tank? Returning from the line to the surge tank preserves the heat already accumulated in the fuel in the surge tank rather than sending it on a long trip back to the nice cool gas tank, and then gives all fuel entering the surge tank a nice long cooling trip up before it gets there....

 

I haven't yet set up a surge tank on a vehicle, but it strikes me as odd that so many people return from the rail to the tank, and then from the tank to the fuel cell, rather than routing main fuel rail return line to the tank for cooling. I suppose it must be logistically more complicated in a way that hasn't occurred to me, since I have only thought about it rather than actually doing it yet.

[Tony D]

"Nice Cool Gas Tank" only applies to after you fuel it from an underground tank.

I have has heated fuel suction cavitation present out here when the tank level drops to 1/4 as the recirculation heats up the tank.

 

Driving across the desert will have your tank at over 140 degrees in NO TIME!

 

I took a Coleman Cooler, sized to hold a full 7# bag of ice, and made a 'cool can' for a 240 we drove up to vegas with...the bag would last about 2 hours of driving, and the water would be over 120F...same as the tank contents! Lower fuel levels just don't work as a heat sink. Even with a full tank, unless you cool the fuel on the way back to the tank it will start heating up.

 

Just depends on the vapor characteristics of your fuel as to when it will cause flash-cavitation on the inlet of the pump. The high-speed electric pumps don't seem any less prone than the diaphragm pumps mounted to the block-even though they are usually much cooler, and nearer the fuel source.

 

The reason to put the return line to the surge tank is when you start sucking air from the tank on your pre-pump, your surge tank still has fuel in it, allowing the main pump to remain flooded...all the fuel returning from the fuel rail keeps that surge tank full during that instant when you are sucking down the main reservior.

 

If you returned it to the tank separately, you would need a much taller vessel, and sized larger in total volume capacity to accomodate the times when you are sucking air.

 

On a track car where you have a single pickup, you could suck air on both right AND left handers, and that would mean a larger and larger fuel capacity in the surge tank for each expected incidence of booster pump starvation.

[z-ya]

The pumps also get quite hot, which I'm sure adds to the heating effect of the fuel. Both tanks get warm BTW.

 

I have a another gauge and a low pressure regulator that I will install tonight. I'll let you know what happens after the next event.

 

Pete

[Nigel]

"Nice Cool Gas Tank" only applies to after you fuel it from an underground tank.

I have has heated fuel suction cavitation present out here when the tank level drops to 1/4 as the recirculation heats up the tank.

 

Driving across the desert will have your tank at over 140 degrees in NO TIME!

 

I took a Coleman Cooler, sized to hold a full 7# bag of ice, and made a 'cool can' for a 240 we drove up to vegas with...the bag would last about 2 hours of driving, and the water would be over 120F...same as the tank contents! Lower fuel levels just don't work as a heat sink. Even with a full tank, unless you cool the fuel on the way back to the tank it will start heating up.

 

Just depends on the vapor characteristics of your fuel as to when it will cause flash-cavitation on the inlet of the pump. The high-speed electric pumps don't seem any less prone than the diaphragm pumps mounted to the block-even though they are usually much cooler, and nearer the fuel source.

 

I'm having a big problem with this right now. When I start out in the morning, after the coolant temp stabilizes, my AFR's are around 13.5 to 1. The longer the engine runs for, and the lower my fuel tank gets, the leaner the AFR. I used to think that it was due to the placement of the air temp sensor in the engine bay resulting in an exaggerated air temp value that would cause the ECU to lean out the mixture. However, I moved the air temp sensor to the outlet of my intercooler, and only saw a slight improvement.

 

It didn't occur to me until recently that it might be a fuel temperature issue. Last night I got to put that theory to the test. I got stuck in stop and go traffic for 40 minutes with a 1/4 tank of gas, and I watched my AFR's climb up to 16.5 to 1! Everything else was constant though (I just realized that I may not have looked at the fuel pressure gauge). The intake air temps matched ambient (30C), and coolant temps remained between 80 and 90C. So, the injectors were still firing for the same duration as they were when the AFR was 13.5 to 1 in the morning, but I was now seeing 16.5 to 1 on the gauge. When I arrived at my destination, I touched the fittings on my sump and they were almost too hot to touch, as was the tank itself.

 

I have a 255lph pump, and with a 1/4 of a tank, that means that the fuel is being completely circulated every 3.5 minutes! There's no way it's going to have a chance to cool down at that rate, since it's doing nothing but pick up heat as it goes through the pump, along the transmission tunnel where it's near the exhaust, through the hot engine bay and back again. I realize now that I inadvertently made thing even worse at the beginning of this year by moving the return line from the top of the tank to the second fitting on my sump. No wonder my poor fuel pump is screaming by the end of a long drive!

 

So, my plan is to insulate the fuel lines in the transmission tunnel, put heat shields around the exhaust, and space permitting, install coolers on both the feed and return lines.

 

Nigel

'73 240ZT

[Daeron]

Tony, THANK YOU. I have been trying to get someone to tell me that for about six months without knowing it. Kinda a "duh" moment that I grabbed onto and wouldn't let go of without the right explanation.

 

I suppose I over-idealized my "cool return line" to the point where the idea made sense.

[Tony D]

Nigel...now you see why I'm such a proponent of not screwing with the EVAP cannister, and keeping it intact, as well. Your example is a good one of something the factory engineers looked at an compensated for: enleanment under increased temperature.

 

What does your fuel do when you heat it? Expand and turn to vapor. The Evap cannister's purge valves will hold several inches of water column in the tank before it 'relieves' pressure to the cannister for collection. While this is mostly an emissions concern after shutdown, it helps keep pump suction head higher running down the road. Unless you are seriously sucking fuel down (like NHRA Fuel Flows) the likelyhood of the fuel getting hot and expanding (and therefore resulting in a pressure rise in the tank) is greater than sucking a vacuum on the thing. This 'heat expansion' will allow for that pressurization on the fuel pump, and combat this kind of issue.

 

What is also happening is 'hot fuel'---you can see this on Diesel Dynos, Detroits were notorious for this, as they gave a specific heat for the fuel when run on the dyno. Simply heat it from 70 to 108 F and you got all sorts of 'efficency increases'... This is a primary reason the Z31 box uses (and I believe all new vehicles currently) a fuel temperature sensor in a lookup table/compensation loop. The BTU content per fluid volume measure decreases dramatically with some of these fuels, and indeed 'cold fuel makes more power' because you get more into the chambe to combust.

 

If you can't touch the fittings without feeling uncomfortable, you are in the 140F range. This is almost universal. See if you can touch your fuel rails and tank bottom after a long run. An Infra-Red Temp gun from AutoZone will open your eyes quite a bit. Once you buy one, you start using it on everything (everybody does it!) and start going "Hmmmm, Oh, Hmmmmm!"

 

We run a 5 gallon cel on the Bonneville Car, have the fuel packed on ice as long as possible, fill the tank and keep the cool can filled with crushed ice to get the fuel as cold as possible. Our fuel rail will sweat at idle on the starting line! It keeps things consistent.

 

I have always suggested that guys with A/C in their car run their fuel line ziptied to their A/C line, and then insulated. You loose some cooling efficiency in the cabin (you will never notice, the pump just runs slightly longer), but you can significantly affect your fuel temperatures everywhere in the system doing this. If it's hot enough for you to consider running A/C, it's hot enough to consider what it's doing to your fuel---works great.

[Six_Shooter]

After reading this thread, it occured to me that this possible phenominon of heated fuel, might be just what happened the other night in my car.

 

I was doing some tuning of the GM EMS, logging, etc, etc.

 

I was out for about an hour, made a few highway passes, and at the beginning, the AFRs were rich, but consistant (my EMS has a WBO2 input). Nearer the end of my session, my fuel was getting low, and the AFRs were going way leaner than I would accept as ok to continue the power tuning I was doing. The car never hicupped, or showed signs of lack of fuel.

 

I later checked the log, between the beginning and end of the session, and compared any concievable parameter in the log that could effect this. everything remained consistant, including injector pulse width, spark was also the same.

 

I had been thinking that maybe my fuel pick-up was being uncovered and not pumping fuel up to the engine, as this only happened when in PE, where the fuel could slosh to the rear of the tank, but again, no hiccup or bucking like I would expect from such a condition.

 

The symptom when away after filling up.

[Tony D]

"I observe things that seemingly have little relevance to anybody but myself."

[cygnusx1]

Here's a thought! Don't circulate a lot of fuel. Design a system to hold Xpsi at the injectors that delivers the amount of fuel needed.

[Nigel]

Here's a thought! Don't circulate a lot of fuel. Design a system to hold Xpsi at the injectors that delivers the amount of fuel needed.

 

Many modern cars use a returnless system, but I believe that's to combat the evaporative emissions issues due to returning hot fuel to the tank. Regardless, I don't see how this would help keep the fuel cooler. If anything, I would think that it would make fuel temps worse since the fuel is trapped. At least if it circulates, it has a chance to be cooled. You'd also end up with a fixed pressure system (not boost referenced), because the regulator would be back at the tank, and a vacuum line that long would be ineffective. Not the end of the world, but it could complicate tuning.

 

I was thinking of something along the lines of a fuel rail cooler. Make a fuel rail with two lines though it, one for the fuel, and the other to pump water through, with it's own separate cooler, and a small water pump. You could use a small reservoir to allow for some expansion that the pump would feed off of.

 

Nigel

[z-ya]

Something like this might help:

 

 

I'll see if the pressure increase in the surge tank fixes the issue before trying to cool the fuel.

 

Pete

[Zmanco]

z-ya, I have a similar set up to you only my surge tank is mounted in the rear right near the 2 fuel pumps. FWIW I don't have any issues with the fuel pressure changing even on hot track days. And I'm still using the stock fuel rail and FPR. The gauge I'm using is a liquid filled unit that I purchased from diyautotune when I bought my megasquirt kit - not sure of the brand.

 

If it's not too much effort, might be worth moving the surge tank out of the engine compartment.

[Tony D]

Actually there is an ECU manufacturer out there now that has fuel temperature and fuel pressure sensors incorporated into the system, and it uses a PWM circuit in the ECU to control the system's fuel pump to allow a returnless fuel line setup to 500HP.

 

Really, if the fuel comes from the tank and does not get heated it should stay cooler overall...longer than if it's recirculating.

 

But once the thermal mass is heated you can't do anything about it. Nigel is spot-on regarding the rationale behind returnless fuel systems: it keeps EVAP Emissions lower because the heat is not vaporizing the fuel and making the emissions.

 

These new systems do use a fuel pressure as well as fuel temperature compensation lookup table to trim pulsewidth to keep btu value delivered to the engine constant. Incidentally, this is how many race sanctioning bodies are monitoring horsepower compliance---when you know spec fuel BTU content, and temperature/pressure/pulsewidth you can by remote telemetry monitor the power the competitors are making...and spot rulebreakers.

 

We are digressing. I'd check to see if boosting the surge tank pressure helps first before we start cooling fuel and going all out in another direction. The easiest way to combat cavitation is to increase suction head---and you can simply do that by pressurizing the inlet to the pump.

 

Now, making a cooling jacket around your surge tank and packing it with dry ice would pay dividends for shorter track sessions.......