Engine behaving strangely and popping hoses off intake manifold

Hey, that looks like the issue!
https://attachments.priuschat.com/attachment-files/2015/10/96212_T-SB-0041-15.pdf

So their repair procedure is to replace the "Purge Vacuum Switching Valve Assembly" (that's this thing right? also known as canister purge valve?) and reprogram the ECU with the "latest calibration" (so like update the software?).

So hmm, I don't see how a malfunctioning EVAP purge valve could cause enough pressure to blow hoses off... I don't think there could be that much pressure in the EVAP system, and even if there was, the gas cap has a pressure relief valve.

But what might make sense is that there's a software bug that's triggered when the purge valve malfunctions… which caused it to do something to the engine that made it pressurize the intake… The ECU can adjust valve timing, right? So maybe it could have not fully closed intake valves before ignition or something like that.

If it is a software bug caused by a faulty purge valve then that would make sense with the repair procedure: replace the faulty valve, and fix the software glitch caused by said valve. If that is true, then, in my case, changing the faulty valve should fix the issue. As long as it functions properly, the software should function properly like it originally did and professional calibration may not be worth the cost. BTW, if the software is updated as part of a recall, would that include the updated calibration that fixes this issue? There's another recall for our vehicle involving a software update (forgot what it was about); maybe that'll fix the (potential) software bug too. Yeah, we need to get that done; keep forgetting...

 

So about that EVAP leak… I made a crappy smoke machine to pump into the EVAP system and saw smoke coming out from the fuel cap area. It couldn't be the fuel cap again, it's brand new and seals tight. So it must be something else near it. I found and removed the panel covering the filler neck in the wheel well and had a look. There's this plastic thing clamped around the filler neck; it snaps together but the snap thing is broken so it is split and loose and I can twist and slide it all around. This is part of the EVAP system, for reclaiming vapors from the filler neck? It was caked with mud so I washed it out. I guess I could just strap it together, but I would want to take it off because dirt and water has probably got inside.

 

So I got that plastic thing out… and it's just some sort of air filter… between that tube and the outside air. There's no hole in the filler neck itself, it's just smooth.

This plastic thing has the metal tube going in to one side of the filter, and the other side is connected to a couple of small holes at the bottom, and a large hole on the inside of the collar.

So besides discovering what this thing is I achieved nothing. Oh well, need to test further.

 

Ok so I believe that is the input of the fresh air line which I reckon is needed to let air in when purging vapors into the engine intake. I did a leak test using the blower end of a vacuum cleaner. Air blew out of that tube.

So looks like the "canister closed valve" has failed to close; either it broke or maybe the rodents finally chewed some wires! Or maybe it's the "fresh air valve" but that doesn't appear to be an active component?

This document is helpful, though it's for 2nd gen so a lot of things are different. https://priuschat.com/attachments/prius-fuel-and-evap-pdf.51790/

 

I tore down the under panels and followed the fresh air tube.
First it goes to a boxy thing in the corner. (77746-47010)

Then another hose comes from that and goes to a square thing plugged into the socket of a cylindrical thing. It also has an electrical connection.

Another hose comes from the other end of the cylindrical thing to a rectangular thing.

The rubber hose in the middle looks like the one that goes to the engine (where I tested from).
The other hose goes somewhere over the gas tank.


I was confused what these things were but after some research I think the first boxy thing is the "charcoal canister filter" (so another filter after the first one up top), then the assembly with the rectangular thing and cylindrical thing together is the actual charcoal canister assembly. It's not clear what each part exactly is but I guess the rectangular thing is the charcoal canister, and the boxy thing attached to the cylindrical thing with the wires connected is the "canister closed valve". do not know what the cylindrical thing itself would be then. I also wonder where the leak tester thing is, the one that makes the mysterious noise a few hours after shut down.

So I guess tomorrow I'll debug what I presume is the valve.

 

Thank you that helps a lot.

Ah ok so that plug-in thing is the leak detection pump.


HMM, the inspection process involving the fresh air inlet and purge port indicates that they are supposed to be pneumatically connected by default. The "purge port" is what I believe goes straight to the engine and is where I was pumping smoke or air from.
Does that mean that the EVAP line in the engine bay upstream from that purge valve is SUPPOSED to be always pneumatically connected to the air inlet filter around the gas cap? So I'm testing it wrong? So then that must mean that this guy was testing it wrong too?

But hmm, if that unit is the leak test pump, the cansiter closed valve must be before it so it can pressurize the system. So it must be part of the unit.

And... according to the 2nd gen document, the canister closed valve is normally open! If that is so then it must be true; the purge line is "normally" connected to fresh air, so you cannot do a leak test there because the air or smoke will just always come out of the gas cap area, making you think it's a bad gas cap!

You know what… I am thinking this EVAP leak is the same issue as what caused hoses to pop off (instead of thinking the pressure blew a hose in EVAP system): a bad purge VSV (next to the intake manifold) as suggested by the TSB for the explosive hose issue. I think it's stuck open now. In that case, while it's open when it's not supposed to be (it's normally closed), it causes a vacuum leak by letting fresh air from the fresh air inlet through the charcoal canister. And of course during a leak test after parking, all the pressure is lost into the intake manifold. But assumedly when the valve is supposed to be open everything is normal as the ECU compensates for the vacuum-leak-like EVAP purging.

I did clamp the purge line though; that should've prevented the P0455. Maybe it didn't seal well enough… or maybe it decided to finalize that DTC anyway, along with P0441 which is expected.

I am going to go straight to that valve and inspect it.
Edit: ah damn it's raining
Edit 2: but that won't stop me

 

Ok so I got the valve out. It wasn't stuck open but repeatedly testing it with a 12v battery revealed that sometimes it would stick open. When stuck, it could be unstuck by a tap, by blowing a large enough volume of air though it, or by reversing the polarity.

I did rinse it thorougly with isopropyl alcohol and tested/dried it on the blower but it still occasionally would stick while the vast majority of the time it worked fine.

My father wanted to get a new one to be sure so I found one and let him buy it but it will take a long time to ship

But since that's ordered I cracked open this valve with a hammer to see what is inside.



Inside there is a magnetic slug on a spring on a plastic pivot piece in the solenoid body. The spring pushes the slug with the seal on top against the top piece, and the solenoid pulls the slug down when powered. There's a broken filter screen in front of the inlet port. The valve parts itself look perfectly fine, though; there really isn't anything that can go wrong. I guess it was sticking because of gunk or fluid. A thorough cleaning (and maybe lubrication) of the solenoid bits would probably prevent it sticking. Unforunately the unit isn't serviceable; you have to smash it open with a hammer and then glue the top back on. But maybe it could be cleaned or lubricated well enough without opening. Maybe a squirt of WD-40 would solve it. Unless it's sticking because of fluid sealing a vacuum behind the slug. In that case you'd have to thoroughly dry all fluids after cleaning it. But whatever, we ordered a new one.

 

Ok so the new valve arrived a week later (Jan 28) and I installed it. Since then there have been no more issues so that marks the problem fixed

So if you are having your engine behaving strangely and popping hoses off, your problem must be that your EVAP purge valve is sticking in the open position. Because the EVAP system is, apparently, normally connected to fresh air, it causes an unmetered intake leak which screws with the ECU. It is in this vacuum leak condition that at some point, somehow, a high positive pressure in the intake manifold blows either the PCV hose or the EVAP purge hose off, creating an even larger intake leak that really screws with the ECU, such that it soon gives up trying to operate the engine, throwing a P3190 and forcing you to pull over.

Who knows how it blows hoses off?? One thought I had is it causes a software glitch in the ECU to adjust things to pressurize the intake. Another thought is a fuel vapor explosion, but I can't see how that could happen. Oh well, at least we know the cause.

So your solution would be to either change the valve with a new one, or maybe try fixing it yourself. It's a simple device and there doesn't appear to be anything to fail; I think it sticks because of residue collected over time that needs to be cleaned out. While it's not serviceable, you can break the top off cleanly by pounding the sides of the seam against a hard surface with a mallot (or maybe a vice would work if you have one). Then you can thoroughly clean all the parts and reassemble. The top fits back on snugly and has a porous edge so I think it would glue back together well with epoxy or something similar.

Thanks to @EBATX for mentioning TSB-0041-15; that was most helpful to pinpoint the exact issue.

Perhaps the past similar issues were also caused by a sticking valve. Maybe there never was a leaking gas cap, and the only reason I felt air blowing out when testing for leaks with a blower was because I tested upstream from the purge valve, and the canister closed valve that is used for the system's own leak testing, is normally open, and the fresh air inlet is right around the gas cap. The previous time when we thought it was a loose gas cap was also odd, but went away at some time after tightening it again which probably might've just been the valve un-sticking with some vibration. The first time I heard the engine behaving strange, went away before the MIL was presumably going to be triggered.

So yeah, if your engine has randomly started to run strangely but then later randomly stopped, it was also probably because the EVAP purge valve stuck open. And it would probably occasionally get stuck again, causing what appears to be any of several different strange issues. Strangest being popping hoses off the intake.