2013~2016 CX-5: How to visually find an underhood fuel leak? HP fuel line reusable?

[hillpc]

Another fuel odor story, this time on a 2015 CX-5 Grand Touring 2.5 liter AWD with about 90K miles. Took the intake manifold off today to try to see where it's leaking.

Background - It's only noticeable after the vehicle is warmed up, when stopped for a red light, or just shut off after driving. I have to turn the fresh air off so it doesn't stink up the cabin if the car is idling. (A problem when I need the defrost on, since fresh air kicks on automatically when I switch to defrost.) It very obviously seems to be coming from under the hood, and nowhere else (this is very likely not a fuel filler pipe issue), but no liquid fuel is ever visible anywhere I can see under the hood or on the ground. I've checked the hoses under the hood, blown through a check valve (steps in the service manual procedure No. 20 for finding a fuel odor cause) but nothing is obvious. The car runs great, with the same great fuel mileage as always, no codes thrown. The vehicle has been absolutely trouble-free in the 5 or 6 years we've owned it.

Today I pulled the intake manifold off to see if I could find any obvious leak spots, since a lot of the potential leak spots are buried under the intake. Some slightly oily areas on the block, somewhat in the area under both ends of the fuel rail. Of course I can't run the engine now. After I got the manifold off, I was able to tighten the high pressure fuel pipe (metal line from high pressure fuel pump to the inlet end of the rail, but only a tiny amount (say 1/12 of a turn). I took off the high pressure fuel pipe and attached a rubber hose to it, to pressurize up the fuel rail with 110 psi air. No bubbles when I painted dish soap solution around all 4 injectors, the connection to the rail, and the pressure sensor at the end.

But I've read that the Skyactiv fuel rail can see over 2200 psi at throttle, so I can't even get close to simulating that, yet. I'd have to use a regulated compressed gas cylinder or hydraulic pump, and make up some kind of (probably?) metal to metal connection - but I don't know what kind of fittings the pipe uses. Can anyone advise?

And is it really true that the fuel pipe cannot be reused? The service manual has conflicting information - "Install a new high pressure fuel pipe because it cannot be reused", and "When removing the high pressure fuel pipe, the end of the high pressure fuel pipe may be damaged because the clearance between the high pressure fuel pump and the high pressure fuel pipe is small. Be very careful when handling the end of the high pressure fuel pipe so as not to damage it."

I'm tempted to just put the whole thing back together and see what I get. Any comments? Chances are very low that I'll take it to any dealership or independent mechanic. Much of my recent reading in various forums about Mazda fuel smells just reinforces this. I've been working deeply on cars as a hobby for about 50 years, and generally refuse to simply throw parts at problems (unless it's cheap and easy).

Thanks,
Pete.

 

[edmaz]

First off, are you completely certain that what you're smelling is raw gasoline? I'm not questioning your DIY capabilities, it's just a question that needs a definitive answer, before proceeding down the diagnostic road.

The HP fuel system is pressurized up to 3K+ PSI, so it's very hard to imagine any HP leak that's not immediately apparent. And the big clue in support of that is the fact that you don't get the smell immediately. There's always at least 1K+ PSI during warm-up, and IMO you would be smelling a HP fuel leak immediately after the engine starts up, even if the HP system was only weeping. Not saying you don't have a fuel leak, only that it's hard to imagine it being from the HP system.

But that aside, and regardless of anything else, I would NOT reuse the HP pipe if it were mine. But as always, your vehicle, your choice.

 

[hillpc]

First off, are you completely certain that what you're smelling is raw gasoline? I'm not questioning your DIY capabilities, it's just a question that needs a definitive answer, before proceeding down the diagnostic road.

The HP fuel system is pressurized up to 3K+ PSI, so it's very hard to imagine any HP leak that's not immediately apparent. And the big clue in support of that is the fact that you don't get the smell immediately. There's always at least 1K+ PSI during warm-up, and IMO you would be smelling a HP fuel leak immediately after the engine starts up, even if the HP system was only weeping. Not saying you don't have a fuel leak, only that it's hard to imagine it being from the HP system.

But that aside, and regardless of anything else, I would NOT reuse the HP pipe if it were mine. But as always, your vehicle, your choice.

Yes, it's unquestionably a raw gasoline smell. I've worked in gas stations as well as on cars and motorcycles, and it's a unique smell.

Regarding the HP fuel pump, what can I make of the attached graphs (oscilloscope traces) about control signals, where it says something about "after warm up"? It seems there's some kind of "spill control valve"; I assume that would be used to dump fuel back to limit pressure. Maybe there's a problem with that system and the pressure stays high and causes a leak at the fuel rail inlet?

I have not put a scope on any signals yet. (Again, not seeing any codes.) I haven't put a pressure gauge on the fuel rail either (not that I have the plumbing to get in there anyway right now).

Regarding pressures (see below), there's some kind of specification about fuel pressure (see below) from 435 to 696 psi, but I don't know under what conditions. I also don't know how many places fuel pressure is measured.

Thanks for the reply. Something to think about. Maybe I should read more on the DTCs (codes, even though I'm not showing any) to learn how the system should operate.

 

[edmaz]

The HPFP is mechanical, driven by the exhaust cam, and so the piston inside of it is always at work trying to build up pressure. The PCM-controlled spill valve is also always working, and is basically a traffic cop. The PCM signals the spill valve continuously - opening the fuel rail side just enough to allow the required amount of fuel into the rail, and at the same time sending the unneeded fuel back to the fuel tank side. And of course this little dance is happening continuously, at potentially a millisecond change rate. There is also a separate relief valve installed in the assembly, which is designed to relieve any over-pressure condition by sending more fuel back to the tank.

The HP fuel system is truly a PITA to diagnose. Due to the high pressure involved in the system, there is no mechanical test tool available to test the (HP) pressure. For the same reason, the HP pressure sensor, which is located in the fuel rail, also cannot be mechanically tested. There aren't very many HP DTCs, although bad spill valve circuits certainly should set some. There are some OBD PIDs available, including HP fuel pressure, however that one uses the HP pressure sensor values. So trying to diagnose the HP fuel system amounts to one big catch-22, and it's basically a part-hanger's delight.

So I don't believe you will get anything helpful from scoping the spill valve signals, but of course I'm not certain about that. And the only way that live OBD data would help is if there are abnormally high HP pressure readings, which would probably mean that the relief valve has failed. You can of course do a pressure test on the in-tank fuel pump, although I can't imagine how that would help you in this particular situation. Also, there's not very much low pressure fuel piping in the engine bay either, and I'm sure you've already checked that for leaks.

Have you considered the possibility of a leak from one of the valves in the EVAP system? Yes, anything like that should set a code and the CEL, so I'm just grasping at straws on that.

 

[hillpc]

Success! Thanks for all the input and the education, edmaz. I was planning to simply put it all back together again (with only new intake gaskets), but you changed my mind, so I also bought a new HP fuel pipe, even though it was a hundred bucks. I didn't want to risk having to take it all apart again.

First time I'd heard the term "part hanger" ("part hanger's delight"). I'll be using that one.

Maybe the leak was so small that not enough had accumulated on the block or head to be odorous enough on startup for it to be smellable.

On the intake gaskets, they were $8 each from the local Mazda dealer, which is cheaper than most all I'd found except for Rockauto, but Rockauto only had 2 in stock. A bunch of places on the internet were charging $20 apiece, and one place had them for 10 bucks plus $25 shipping (each)!

I tested the car tonight, getting it good and warmed up on at least 20 minutes of driving, with a decent amount of full throttle blasts to make sure the fuel pressure would get up high. No trace of gasoline odor.

I conclude that the fuel pipe had become slightly loose (since I was able to crank it down about 1/12 of a turn before unhooking it, or the pipe was compromised in some manner.

Thanks again!

Pete.

 

[edmaz]

Happy to read that you were able to resolve the problem, and we know how great it feels to have DIY success like this! Based on what you wrote I agree that something in the HP line must have become just a bit undone, and also believe that you made a wise decision and a good investment spending that extra $100.

The follow up information that you posted may very well help others having the same issue in future, and it's always appreciated when a problem is closed out the way you did in this thread