Poor warm starts & fuel odor

[FrankAZ]

Hello. Long term reader; first time poster.

My wife drives a 2004 Volvo V70 2.5T FWD we have owned from new. Loves it, refuses to swap it, so we're dumping money into it until something big happens.

Recently it has suffered some rough idling and slow starts. I think I have been logical in my diagnosis and my own repairs but something screwy is going on. Open to suggestions.

To begin with the idle was rough and it would dip and splutter, especially when cold. Sometimes when driving I could feel an occasional misfire. So, aided by a thrown code (cyl 5 misfire) and mindful of the vehicle age (155k miles) and time since last pertinent maintenance I replaced the spark plugs (properly gapped), coils, and air filter. Immediate improvement and the car ran smoothly. They certainly needed replacing.

However, once the misfire was cured I still had a slow start, especially if the car was warm. Starting from cold in the mornings was OK; starting within a few minutes of turning off was OK. Worst was trying to start about 1 hr after turning off. The engine cranked, would catch, splutter, surge, then settle down. There was a tell-tale small of fuel around the exhaust. Classic flooded engine behavior.

Ah ha, I thought. A leaking injector! So, I replaced all 5 and things somewhat improved but not completely. There is still a slow warm start accompanied by an unburned fuel smell.

So then I looked at the long-term fuel trim and other OBD parameters. The engine runs rich (LTTRM -(15-18)%, repeatable after clearing). In succession I substituted the MAF with known good, replaced the Fuel Pressure sender, replaced the Boost Pressure / Inlet Air Temp (which seemed to be reporting ~15F low), and along the way also the Vacuum Booster switch which I noticed broken, though this would not introduce a significant vacuum leak. No improvements. The OBD reports the same pressure as an external diagnostic fuel pressure gauge (400 kPa / 58psi at run; 470kPa / 68psi start). I do notice that the fuel rail pressure drops quite quickly (~10psi per minute) after I stop the engine so there is a leak somewhere, back past the check-valve in the pump or through an injector.

For good measure, the O2 sensor is behaving as expected. Moves between 0.010 and 0.750 volts as I would expect with acceleration, deceleration, cruise. Nothing untoward in those readings.

So, after replacing or experimentally substituting everything except the fuel pump/filter/check valve assembly (which is difficult to get to in the fuel tank) I still have all the symptoms of a leaking injector (or two, or three) even though they are new (not refurbished, new).

Symptoms:

• Slow starts after ~ 1hr symptomatic of a flooded engine. Starts more easily full-throttle; fuel smells.

• LTTRM values strongly -ve (-16%) - sometimes triggers a bank 1 running rich code.

• STTRM values swing +/- as expected with engine speed, load, ...

• O2 sensor seemingly working, see also expected MPG note.

• Fuel rail pressure is OK at start and run; but reduces quickly after engine cut.

• MAF values credible and same as with a known-good unit. Also consistent (corrected for rpm, cyl#, & CC) with my own car's MAF.

• Onboard fuel pressure measurements in line with diagnosis pressure gauge.

• Vehicle performance does not seem unusual.

• Very black (not grey) soot/residue in tail pipe indicative of running rich

• Engine oil seems OK. No obvious signs of fuel dilution. Smells OK, and (shame) tastes typical.

• New boost pressure / IAT sensor still reports IAT ~ 15-20F higher than ambient, but the air has a long journey around a hot engine before it reaches the sensor.

• Despite all of this the MPG seem about normal so the fuel mix in the engine at run is ~OK after being compensated for.

I think I still have bad injector(s) but I am loath to go through the hassle of a return/replace without more confidence. With the Volvo 2.5T engine the fuel supply pipe disconnects when the fuel rail is removed so I am unable to eyeball the spray patterns from each injector or check for leaks under pressure. I tried jury-rigging an injector test using an air compressor but could not clearly determine any one or two injectors behaving differently than the others (of my 10 available) so I doubt the validity of my test rig.


SO: Is there anything I have missed? Without wanting to guide group-think too much, does it still look like a leaking or over-delivering injector, or two?

Thanks for suggestions.

Frank

[RickHaleParker]

Did you take it out on a long run and give the ECU time to learn the new patterns and retrain?

The software in the Engine Control Unit is Artificial Intelligence. It make decisions based on past mistakes. Give it time to learn from new mistakes.

[FrankAZ]

Thanks for the suggestion.

While the V70 is under my care I have a 20 mile trip each morning, the first 10 miles ~45-50 MPH with occasional traffic stops; the next 8 miles on an open highway ~75 MPH. If I reset values before I leave the house (clear codes, start from cold which is easy after an overnight engine stop) then the LTTRM value converges to -16% - -18% on the journey. I.e. it is adapting to lean-out an otherwise nominally rich mix. If I leave it alone the LTTRM stays at about that value +/- 1% over several days.

Is that the training? And is it enough? It doesn't change of much less improve with longer training.

I agree that the ECU is adaptive to natural variations in conditions and the system. Reading up on it informed me that LTTRM was that adaption and that a normal stable value would be +/- 10% to accommodate variations in injectors and sensors. For reference, my other car has a LTTRM of +0.72%.

My concern is that the ECU does seem to adapt, but perilously close to the check-engine threshold of +/-25% which is occasionally exceeded when a few warm starts are thrown in. That excursion causes a 'Bank 1 running rich' code. That behavior is leading me to the conclusion that I still have bad injector(s) since on the 2.5T engine the injectors are the only way fuel is introduced to the engine. (No cold start injector) and almost every other fault would tend to cause a running lean condition. And, the O2 sensor seems to be successfully driving cycling while in closed-loop to control the 3-way catalytic converter efficiency and generally maintain stochiometry at full adaptation.

Thanks,

Frank.

[abscate]

Have you evaluated engine coolant temperature sensor performance? That's s. I'm on cause of rich running.

[FrankAZ]

Thanks.

I considered that too, early on, but put the thought aside given the issue with actual and reported warm starts. The car knows the engine is at the right temperature as far as I can tell.

The engine seems to be running at nominal temperatures. The dash gauge shows the same movement as always and the OBD reports 194F when fully warm. That could be a busted sensor, but it does show a plausible temperature at other times rising from near ambient after an overnight. Plus, the coolant pipes seem plenty hot, and if I torture myself with full-on cabin heat (I'm in AZ, so aargh!) the dash gauge and OBD take a mite longer to rise to nominal so I think the thermostat valve is good. The coolant expansion tank vents some pressure as it always did if I crack the cap open after a run but does not do anything as hazardous as splashing hot vapor.

In short, I am not seeing anything which would suggest that the engine temperature is being misreported, is off, or is affecting the warm start.

But, thanks for the suggestion. An ECU thinking the engine were cold and giving it some choke (old school, or longer injector triggers nowadays) to enrich the mix would cause similar symptoms.

My car seems to be *really* flooded though. I'm imagining an injector leak causing a good dump of fuel into the inlet manifold over the course of 30-60 minutes and really saturating all the air in the manifold and back up past the turbo to the air filter. My long start is the engine having to suck all that enriched fuel/air through the inlet system before a combustible mix arrives.

During a short stop of course the injector does not have time to leak a significant amount of fuel and charge pipe is not far off pure air; a long stop allows the charge pipe to vent gasoline vapor back through the air filter and to atmosphere. Hence the faint fuel smell around the car in the evenings, and the heavier fuel smell from the exhaust after starting the engine after an hour or so.

This dialogue is useful. I am convincing myself to buy another set of Bosch injectors from a reputable supplier and argue for the return of the first set if the new ones solve the issue. (My first set of replacement injectors were not OEM.... may be a lesson learned... watch out for a scathing Amazon review if that is the case).

Frank.

[jonesg]

I would pressurized fuel system with key on, then leave it to leak. Crank starter with ignition didabled, then check plugs/cyls for flooding.

[FrankAZ]

Hello.

Problem solved.

I replaced the garbage injectors (DOICOO) with OEM Bosch injectors. The ones I should have bought in the first place before I tried to go cheap.

My LTTRM (80 miles) is now -0.80% and there are no more starting problems.

Thanks for everyone's suggestions. Now I know not to trust cheap injectors.

Frank.