2006 MkII Skoda intermittant engine faults

Good evening folks,

Apologies for what will most likely be a long (first) post - I've been grappling with this problem on a part-time basis for about a year or two now and have trawled these fora extensively for ideas and inspiration (thank you to all contributors). But, whilst invaluable, the problems I'm observing still remain. I'll structure this post with (i) a description of what happens, (ii) what VCDS scans reveal and (iii) what I've tried so far to no avail along with questions I'm grappling with. I'm hoping that I've overlooked the blindingly obvious somewhere and some wise person here will nod sagely and say "ah yes, this is a common fault and is solved by X". 

Firstly, the car in question is a 2006 Mk II Skoda Octavia 1.9 TDi with ~178k miles on the clock with chassis type 1Z (1K0). The car lies idle for between 4 and 8 months at a stretch and is then used intensively in a mountainous environment (not UK based). The battery is removed when the vehicle is laid up and has been renewed a number of times. Currently, the battery is brand new and fully charged. The oil level is normal, and the car has recently (within the last year) had an oil and filter change. The only warning light that is persistant is due to an EGR fault (see section 2). There have been turbo boost problems due to coking, but extensive use of turbo/injector cleaners have solved this problem (drives perfectly up mountain roads / autoroutes and on VCDS the turbo set point pressure is now tracked closely by the measured pressure).

(1) Observed faults

After an idle period, the vehicle is very hard to start. Upon turning over, the engine attempts to fire and usually runs for two or three revolutions before cutting out. It then takes up to two minutes of cranking to start. It's almost as if the fuel supply is compromised (see section 3) or there's air in the injectors. Once started the vehicle then usually runs smoothly. However, after about between 5 and 10 miles of driving (faultless performance) the traction control warning light starts flashing. This is followed, in swift succession, by the rev counter dropping to zero, the engine fault warning light [the glowplug warning light] illuminating and then the red flashing oil warning light illuminating along with an audible alarm sounding. The first time this happens, the driver usually wets themselves, pulls over and turns the engine off. After a few seconds wait and a change of pants, the engine restarts perfectly and - 9 times out of 10 - runs smoothly with all lights bar the EGR warning light extinguinshed. 1 time out of 10, the engine restarts and runs very roughly: this clears after about 10 minutes. The warning light son-et-lumière routine usually then repeats another two or three times over the course of 10 to 15 miles - at this point dipping the clutch, turning off, waiting 5 seconds, restarting and continuing to drive clears the fault. No change of pants needed. And then the car is mostly fine with up to 1000 miles driven over the course of a few weeks without any further drama.

Mostly fine means that, on occasion, the starting problems resurface and getting it out of the garage becomes a cranking-fest once again. But the warning light son-et-lumière routine does not manifest again if this does happen.

There is no pattern I've noticed linking weather conditions to starting problems. The problems are essentially consistent whether it's sub-zero or in excess of forty degrees outside (no exageration, temperatures measured in degrees C). Car is kept in a standalone concrete cave:it can be a bit damp, but not excessively so: definitely better than being kept outside.

(2) Known issues reported via VCDS

The car has been VCDS scanned within an inch of its remaining natural life. Known issues are:

• knackered glowplugs on all cylinders (but starting in hot weather is the same as in cold weather- either first flick of the key or a crank-fest);
• EGR valve fault (due to be replaced - I've removed and decoked the original valve, but the diaphragm is probably knackered. Vaccum supply to EGR feels fine and EGR position remains the same whether or not vacuum is applied);
• occasionally a crank sensor fault appears - but inconsistently and usually at a significant number of km/miles before any observed problems;
• HVAC refrigerent low pressure (functional A/C would be nice on the 40+ degree days, but open windows work just fine).

(3) Things I've tried to no avail

First thought was that excessive cranking was substantially draining even a new battery and that all the sensor voltages were on the lower border of acceptability as a result. Sometimes the car is driven "gently", at low rpm in high gears, hence the charge rate from the alternator probably isn't that great. Add that on to drainage from the electric power steering worked hard on twisty roads, plus full headlights, made me wonder whether the number of joules extracted from the battery was less than that going in, hence resulting in the dashboard Christmas tree after 5 to 10 miles. This theory is now largely debunked after no discernable difference in the problem when driven "gently" and in a more, ahem, "spirited" manner... The car has also had two new batteries - again with no discernable difference in how the fault presents. Moreover, no battery charge warning lights appear.

Second thought  was that it was a fuel supply issue after being left idle with, possibly, a non-return valve leading to fuel lines back-draining into the tank. Strategic removal of fuel lines from the fuel filter and cranking, however, resulted in various fountains of diesel being sprayed on the car cave's concrete walls and no obvious bubbling due to trapped air. Thank goodness for decent safety specs and nitrile gloves...

Third thought was that there was a genuine oil supply problem (despite no seizure - something which I do have first-hand experience of in a Mk1 Occy). Everything's fine on that front - oil levels normal, no drips, no leaks, no spray of oil in those hard-to-see places buried deep in the engine bay. I haven't been able to measure oil pressure with a separate pressure gauge, but given that the red flashing oil warning light is also accompanied by other strangeness such as zero rev counter plus traction control warnings, makes me think that this is a complete red herring.

So my current thoughts are needing some guidance. If it is indeed a crank sensor problem, then why can I drive the car for up to 1000 miles without it missing a beat, and with problems generally coming only when the car is (attempted to be) started? If it was a glowplug fault, then (i) why does it happen when it's hot enough to melt tarmac in the shade and (ii) why the dashboard Christmas tree? If this is in some way linked to the EGR fault, why the intermittancy when the EGR is consistently seized shut and why the initial dashboard Christmas tree? If, Valhalla forbid, it's an ECU fault, then again why can the car be driven faultlessly (and spiritedly) for hundreds of miles once the initial problems have dissipated?

If anyone has any ideas, or has experienced something similar, then I'm all ears. FWIW I'm a(n) (professional) engineer, so I really do want to understand and mentally sanity check the chain of events and the various causes / effects before committing the expected blood/sweat/spinal traction/credit-card-ruiniation sacrifices to effect repairs. 

Thank you for reading this far (cocoa laced with rum to help you sleep may be better) and thank you in advance for any ideas!

Edited 11 October, 20241 yr by crashedOut