Grand Caravan Seems to Flood When Hot, Won't start for 20 minutes

Maintenance/Repairs
,
41

This vehicle has a running fuel economy average of some sort displayed on the dash. It is not a chart, as I had with a 2001 Prius. Just one number “X mpg”. Mine, during periods when I am 95% city driving, read 9.3 to 9.5 mpg.

I haven’t been keeping a precise ledger; I typically refill when gauge is “below empty.” I then divide the number of gallons I put in by the number of miles driven at the time of the next refill. I come up with numbers around 10 mpg. Except when i do a lot of highway miles; then it is much better.

Incidentally, I just looked it up and the tank is a 20.5 gallon tank, not a 20 gallon tank.

Refilling when “below empty” typically means putting 17.5 to 20 gallons in, for me.

i just don’t know enough about the “weighting” of the mpg figure on the dash to say more about its precise accuracy. But I am getting around 10 mpg when only driving city.

Is knowing whether the number on the dash is perfectly accurate or not going to tell us something here?

| Nevada_545
November 14 |

  • | - |

foggbooks_182561:

Yes. I go whole tanks that are 95% city. And I need to refill 20 gallons after about 200 to 220 miles. So, yes, it is getting terrible mileage.

This vehicle has a 20-gallon fuel tank, refilling when “empty” would be 15 to 18 gallons.

foggbooks_182561:

I get the impression that the mpg calculation is referential to some recent period of driving, but I don’t know exactly how it is calculated…

If you have proven your fuel economy matches the vehicles display, you either understand basic mathematics or entered the miles/gallons into a calculator. How did you calculate the fuel economy?

42

… and thank you for your reply.

This vehicle has a running fuel economy average of some sort displayed on the dash. It is not a chart, as I had with a 2001 Prius. Just one number “X mpg”. Mine, during periods when I am 95% city driving, read 9.3 to 9.5 mpg.

I haven’t been keeping a precise ledger; I typically refill when gauge is “below empty.” I then divide the number of gallons I put in by the number of miles driven at the time of the next refill. I come up with numbers around 10 mpg. Except when i do a lot of highway miles; then it is much better.

Incidentally, I just looked it up and the tank is a 20.5 gallon tank, not a 20 gallon tank.

Refilling when “below empty” typically means putting 17.5 to 20 gallons in, for me.

i just don’t know enough about the “weighting” of the mpg figure on the dash to say more about its precise accuracy. But I am getting around 10 mpg when only driving city.

Is knowing whether the number on the dash is perfectly accurate or not going to tell us something here?

| Nevada_545
November 14 |

  • | - |

foggbooks_182561:

Yes. I go whole tanks that are 95% city. And I need to refill 20 gallons after about 200 to 220 miles. So, yes, it is getting terrible mileage.

This vehicle has a 20-gallon fuel tank, refilling when “empty” would be 15 to 18 gallons.

foggbooks_182561:

I get the impression that the mpg calculation is referential to some recent period of driving, but I don’t know exactly how it is calculated…

If you have proven your fuel economy matches the vehicles display, you either understand basic mathematics or entered the miles/gallons into a calculator. How did you calculate the fuel economy?

43

Thank you for your comment. I am interested to understand the above paragraph better. “Residual pressure was low”; that would refer to a leaking regulator or pump that lets the fuel pressure off during the 10 minute incubation period for the problem? That is my understanding of what you mean here; please correct me if i misunderstand. So if that, then the “in tank or near tank high pressure fuel pump’s 40 to 60 psi will pus the vapor…” but what if the fuel pump is failing to generate that much pressure? What do others think of this idea? I don’t know if the Dodge Grand Caravan 2010 3.8L is “returnless”… How long would it take to “purge” the vapor through the injectors? I’ve tried cranking and cranking, but it doesn’t start. Ya gotta wait 15 to 20 minutes after the lock (or whatever it is) for the van to start.

44

Residual pressure low is, indeed, when the pump’s anti backflow valve or injectors are leaking so the pressure reads low before you turn the key.

Vapor lock occurs on carbureted cars with pumps that only deliver 3-4 psi that can’t push the vapor bubble past the carb’s float valve. At pressures that low, your van won’t start or run, hot or cold.

Any vapor in the fuel lines of an injected engine would be purged in one or 2 cranks and the engine should start on a return or returnless system. If the pump’s pressure is too low it would not start hot or cold.

Testing the pressure would end the speculation and provided actual data so go do that and post the results.

45

From the FreeAutoMechanic article:

Vapor lock is just air trapped inside the fuel system. Simply removing the gas cap would releive the situation if this is what is happening.

This is just wrong. That is not what vapor lock is. That is cavitation.

From Axle Addict:

Although common in older vehicle models, it wasn’t until the appearance of the modern fuel injection system that vapor lock became more prevalent in carbureted engines.

That is not just wrong, it makes no sense.

Fuel-injected engines use a more volatile fuel that is easier to vaporize and mix with air for a much better combustion. So this new fuel has a lower boiling point.

This is also wrong and makes no sense. During the days we had both FI and carb’ed cars the same fuel was used for both. It was also at a time before the introduction of ethanol.

The Grabner Instruments white paper was interesting but they did not say how vapor-lock might occur in an FI engine. The most interesting comment was the 5-6 bar testing request from manufacturers for vapor pressure. 5 Bar is 73 psi and 6 is 88 psi, both more than the fuel pressure supplied by most non-direct injection engines. If the crop of modern direct injection engines in-tank boost pumps supply lower pressure than 40-60 psi, that could vapor lock and starve the high pressure direct injection pump.

The system the OP has is not direct injected, however.

1 Like
46

Tester

1 Like
47

Why are you doing that ? The fuel in your tank acts as a cooling agent for the fuel pump . While it may not damage the fuel pump why take the chance . Plus the fact that a power outage might cause the gas stations to close .

48

It appears to have been written in 2009. Look in the document properties…

I am also concerned any time I read any kind of article like this written by a company in the business of supplying tools or equipment related to the subject matter.

1 Like
49

The video (even if sponsored by Delphi :grimacing:) seems to reinforce my comments about DI and about non-DI fuel injection. The DI systems dead-head against the high pressure pump and have nowhere to bleed off any vapor created by heat-soak. Hence the software work-around.

But the OP’s vehicle still has non-DI fuel injection system…

50

Never experienced vapor lock on any fuel injected car I’ve owned. But It seems like if the fuel tank to intake manifold configuration is tank/pump/fuel line/fuel rail/pressure regulator/excess back to tank, about the only place vapor lock could occur is inside the pump. Until a fuel pressure and hold test is done, guessing is about the only path forward.

Carb’d cars in days of yore, vapor lock more common. First b/c fuel pump is located midway from tank to carb, so is pulling fuel rather than pushing much of the way. And fuel pressures (i.e. forces) are lower.

OP, suggest to ask a shop for a quote for fuel-rail pressure/hold test, might not be as expensive as you think. If you want/need to do the test yourself, and need to T-in b/c of your Caravan’s config, I had to use that method w/my prior VW Rabbit. Finding the appropriate adapters is the hardest part. Fuel injection pressures are very high, suggest to only use braided fuel hose. And have a big fire extinguisher on hand. Be sure to first remove residual pressure before breaking into fuel line. I’ve done that myself by disabling fuel pump, then starting engine and letting it run until it stops b/c of lack of fuel.

Make sure you really need to t-in though. No way other than that method on Rabbit, but on Corolla I can remove the aux fuel injector (cold start injector) and easily access the fuel rail pressure there.

51

I had seen this also (Delphi video); thanks for adding it to the thread here. Does this apply to my vehicle? Mustangman says I don’t have Direct Injection.

Is this just a reference for your debate with Mustangman, or is it directly relevant to my situation?

The Delphi guy talks about a “hot soak”; what exactly does he mean? Does he mean a hot engine sitting on a hot day? But my problem has occurred recently even at 50is degrees in the early evening… so not a “hot day” situation.

On the other hand, if a “hot soak” just means the engine is sitting awhile, after having been warmed up, then this would apply to me in that regared (but apparently not in regard to the direct injection issue).

Thank you all for your continued thoughts. This is an even bigger mystery than I realized. It seems as if none of us even have a theory that fits the symptoms… ?

| Tester
November 14 |

  • | - |

https://www.youtube.com/embed/GCaeTNVOvRI

Tester

52

I just didn’t realize that it puts stress on the fuel pump, until now. I considered it a time-saving maneuver; also it makes tracking fuel economy in one’s head easier. As for power outages, never thought much about that.

| VOLVO-V70
November 14 |

  • | - |

foggbooks_182561:

I haven’t been keeping a precise ledger; I typically refill when gauge is “below empty.”

Why are you doing that ? The fuel in your tank acts as a cooling agent for the fuel pump . While it may not damage the fuel pump why take the chance . Plus the fact that a power outage might cause the gas stations to close .

53

Here is another overall question that may help in our effort to logically determine what is going on.

If the fuel pump is weak, and thereby causing the P0171 “lean condition,” then WHY do I ONLY get the P0171 condition, referencing only one side of the engine ("System Too Lean: Bank 1). If all the fuel is low pressure, wouldn’t I get an identical message for “Bank 2”? That would be P0174 ("System Too Lean: Bank 2). I have NEVER got that message.

Doesn’t this tell us that it has to be something particular to “Bank 1”? Rather than something that would affect both sides of the engine?

Like, perhaps, a leaking injector? Or what else?

Any thoughts?

I just didn’t realize that it puts stress on the fuel pump, until now. I considered it a time-saving maneuver; also it makes tracking fuel economy in one’s head easier. As for power outages, never thought much about that.

| VOLVO-V70
November 14 |

  • | - |

foggbooks_182561:

I haven’t been keeping a precise ledger; I typically refill when gauge is “below empty.”

Why are you doing that ? The fuel in your tank acts as a cooling agent for the fuel pump . While it may not damage the fuel pump why take the chance . Plus the fact that a power outage might cause the gas stations to close .

54

Just to "prove’ that there is no test valve for the fifth generation dodge caravan, here is one of several posts that I have seen that confirm this:

https://www.dodgetalk.com/threads/where-is-the-fuel-pressure-test-port-located-on-the-2012-dodge-grand-caravan.662730/

Why would they make it so hard to test fuel pressure?

Here is another overall question that may help in our effort to logically determine what is going on.

If the fuel pump is weak, and thereby causing the P0171 “lean condition,” then WHY do I ONLY get the P0171 condition, referencing only one side of the engine ("System Too Lean: Bank 1). If all the fuel is low pressure, wouldn’t I get an identical message for “Bank 2”? That would be P0174 ("System Too Lean: Bank 2). I have NEVER got that message.

Doesn’t this tell us that it has to be something particular to “Bank 1”? Rather than something that would affect both sides of the engine?

Like, perhaps, a leaking injector? Or what else?

Any thoughts?

I just didn’t realize that it puts stress on the fuel pump, until now. I considered it a time-saving maneuver; also it makes tracking fuel economy in one’s head easier. As for power outages, never thought much about that.

| VOLVO-V70
November 14 |

  • | - |

foggbooks_182561:

I haven’t been keeping a precise ledger; I typically refill when gauge is “below empty.”

Why are you doing that ? The fuel in your tank acts as a cooling agent for the fuel pump . While it may not damage the fuel pump why take the chance . Plus the fact that a power outage might cause the gas stations to close .

55

That’s often a phrase used for describing the condition when a car problem appears. Some sort of heat related problem in other words. For example if you said the engine cranks ok but won’t start with a hot soak, then the problem is likely caused by something heating up and not working b/c of the heat; e.g. crank position sensor, fuel pumps are sometimes a failure items with hot soak.

56

A fully warmed up engine, when stopped, retains a lot of heat and this heat builds up under the hood, so some components get hotter than they would with a running engine that has the cooling system running and some air circulation under the hood.

1 Like
57

So today I “rented” a pretty good fuel pressure test kit from O’Reilly’s (apparently, the “rental” is free if returned within 48 hours, but after that, you bought it). It had the necessary “T” fitting to add the tester to the “jumper” fuel line (or rather to a replacement “jumper” fuel line) that I pieced together from generic rubber fuel line.

The fuel pressure tested at 36 psi. From a posting on the Chrysler Forum I find that it should be 58 psi, give or take 5 psi, so 53 to 63 psi. On the “plus” side, it only lost about 8 psi after 5 minutes, after engine turned off… so theoretically that is barely within tolerance for loss of pressure, I guess. Again, got this info from another forum, so I hope the poster was correctly quoting “the manual.”

So it seems that this is definitive proof that the fuel pump is failing. I can’t think of anything else that could cause low fuel pressure (unless, perhaps the fuel regulator, which, for this vehicle, is integrated into the fuel pump, though it c a n be separately replaced. I read somewhere that the price of the regulator only is about $200 less than the whole pump. Seems to me I should just replace the whole pump, due to labor cost of dropping the tank and such… but I am open to alternate advice.)

I will probably get the fuel pump replaced at a shop; I have an appointment for Thursday. But I do wonder whether m a y b e I should attempt it myself. It is a somewhat daunting prospect, lacking a lift. Does anyone know what the most difficult parts of the job would be? Have you done it?

Thank you to Tester for encouraging this test as the next move, over a month ago.

It seems this may explain all the symptoms, esp. if lower fuel pressure makes vapor lock more likely. I guess it would, but I am still not entirely sure. Also, I still don’t see why P0174 didn’t show up as well as P0171…

That’s the latest. Thank you to you all for all your informative comments. I have learned a lot.

| shanonia
November 16 |

  • | - |

A fully warmed up engine, when stopped, retains a lot of heat and this heat builds up under the hood, so some components get hotter than they would with a running engine that has the cooling system running and some air circulation under the hood.

58

I’ve removed many gas tanks over the years for fuel pump replacement, and the most difficult part of the job even for professionals is removing rusted gas tank strap bolts without snapping them off.

image
image1500×2000 328 KB

Tester

59

That might not be a big problem here in Portland, OR, given we are don’t have a lot of snow and concomittent salting of roads, nor are we next to the ocean. The van seems to be pretty well preserved, so I would hope that the strap bolts are not badly rusted. Does one just soak them with PB Blaster or something? Is it hard to get it to penetrate “up” from underneath? When you say “even for professionals” I am a bit worried… are there any “tricks” to it?

Thanks again for staying with this thread all these weeks, Tester.

| Tester
November 16 |

  • | - |

foggbooks_182561:

Does anyone know what the most difficult parts of the job would be?

I’ve removed many gas tanks over the years for fuel pump replacement, and the most difficult part of the job even for professionals is removing rusted gas tank strap bolts without snapping them off.

Tester

60

From what I see, for a 2010 GC w/3.8L engine, fuel pressure spec is 59 psi +/- 5 psi. Test is done with ignition “on” but engine not running, scan tool is used to command fuel pump on via fuel pump relay. Fuel pressure hold spec appears to be no more than 10 psi drop in 5 minutes.

When I tested the fuel pressure on my VW Rabbit years ago, using similar methods, I had to bleed the air from the lines after getting everything connected, otherwise I’d get faulty readings. Hopefully your testing method addressed that requirement.