Diesel Vectra 2 litre 2004 - starting problem

the one way valves are an add on, trying to cure the starting fault.

to test whether the injector seals are leaking, apply vaccum to each spill pipe, if oil comes out the injector seals are faulty or the head is cracked, overtightening new seals breaks them too. each one should hold a good vacuum.

Mrcheerful wrote: [snip]

engine oil or fuel oil?

If fluid comes out which is black it will be engine oil, the ideal is that nothing comes out and vacuum holds. Check each one in turn, it may be that only one is faulty or all of them. If the injector seals leak then air and engine oil enter the fuel system, just the same as if the spill pipes are leaking, difficult starting after standing is the result.

Which takes us back to my original question about the connection between the injector and the lubrication oil system - where is it?

I'm trying to visualise the comparision to a petrol engine, where the cylinder head has a threaded hole for the spark plug, and the plug has a copper sealing ring so that combustion products don't leak out past the thread. But this hole is drilled through solid metal - it doesn't go through any oilways or water chambers.

In a diesel engine the injector goes in what amounts to the same hole - yes?

If the head is cracked you would see oil in the coolant and the engine would deteriorate rapidly - but other than this starting problem it's run fine for the past year, about 20,000 miles.

The Haynes manual describes the procedure for priming the injection pump, for occasions when the inlet pipe is empty because for example the fuel filter (or injector pump itself) has been changed.

The procedure uses an electric pump connected to the return outlet of the injector pump which draws fuel from the tank, through the filter and the injection pump, thereby priming it. You then crank the engine, so the injector pump fills the pipes from itself to the injectors, and it should start fairly quickly. Once it's been running for a few minutes you stop it, disconnect the priming pump, and reconnect the return pipe back to the tank.

So I imagine if you use a vacuum to suck at the spill pipe outlet from an injector, this should suck clean fuel from the tank through the filter, injection pump, and the injector itself - so you would expect to see clean fuel coming out. Is that correct?

there should only be fuel come through the spill pipe with the injector operating. the injectors on that engine live inside the head, so are surrounded by engine oil. an interesting test for air entering would be to clamp the spill pipe before it goes back into the injector pump and then turning off the engine, if it starts ok next day then it should show if air being pulled in that way is the problem.

Mrcheerful wrote: [snip]

But presumably the injector is built inside a tube or housing of some sort - all metal - so the engine oil is outside that tube. There should be no path from the oilway to the interior of the injector, surely?

Or is there a hole in the housing (perhaps there for ease of manufacure) which must be covered by a seal?

The engine always starts OK the next day, and the day after. It has to be left for 3 days before it has problems starting. Typically it is used on a Friday night but not over a weekend, so won't start on a Monday morning.

I'm tempted to fit transparent tube in place of the black rubberised stuff between the filter and the injector pump inlet, so I can see if an air bubble forms.

if it formed there then it would likely show that the fault was at the filter (also common, both seal and actual housing). you could try putting a clear section on the spill pipe connection to the fuel inlet before the pump, that would show if air was being drawn in on that circuit, likewise it would show black oil.

An update for you:

There were two non-return valves fitted in the fuel lines; one before the fuel enters the filter, the other between the injector pump return and the pipework leading back to the tank.

We removed the latter, from the pump pack to the tank. Starting after being left over the weekend still a problem.

We removed the former, noting that the short lenght of black rubberised pipe connecting it to the pipe leading to the filter did seem to be very brittle, and might have been letting in air. So replaced this pipe with new. A day later (so before the problem would have shown itself) I replaced this with a 4-inch length of transparent pipe (a piece of sight tube for a CH oil tank, so designed to tolerate paraffin). Once the fuel had pumped through this showed only very tiny bubbles - pinhead sized - after standing a few hours.

Left it in driveway on steep slope, nose up, from Friday night to Monday morning. Still very tiny bubbles in sight tube. Engine started immediately!

We don't know why the valves had been fitted, but it is possible they were a work-around because the injector seals had been leaking. The dealer I bought the vehicle from in November 2012 had these seals replaced under warranty. I talked to the mechanic who did the replacement and his diagnostic work did seem credible, so I've no reason to doubt that it was done. Clearly the work-around brought its own problems ...!!!

hooray.