Select a technical image to view it in detail.
The owner of an N53 engine came to me with a problem. The engine was properly maintained, all systems worked, and it ran in stratified charge mode. Then it suddenly started to “faint” on the highway. The EML (Check Engine) symbol lit up in the KOMBI, and one or even two cylinders were switched off.
The owner stopped at the roadside and restarted the engine as quickly as possible. Surprisingly, the engine recovered normal performance, and even the EML went out.
What the DME fault memory showed
A little later, the owner checked the fault list in the DME memory and found this:

All cylinders of the second bank had been switched off, evidently one after another. A range of Lambda probe faults was also stored.
Another entry interested me more, and I marked it in the image:

It shows that shortly before the failure, the injector of cylinder 6 began to leak heavily. At idle, the DME tried to even out the cylinders’ smooth running. It reduced the planned injected fuel quantity by 3 mg/stk, but without success. The leak was really severe, and if an injector starts or keeps leaking heavily while driving, the consequences can be serious.
How I think the failure developed
- The injector of cylinder 6 started to leak heavily, as the fault entry indicates.
- The DME switched from stratified charge to homogeneous mode, which is its normal response in such cases.
- The mixture in cylinders 4 and 5 became lean and misfires began there. The spark plug of cylinder 6 may also have been flooded, causing misfires in that cylinder too.
- When the DME detects misfires in the cylinders of the second bank, it switches them off, then tries to restore them after a moment.
- Because of the leaking injector, the Lambda probes of the second bank reported completely implausible values, and the DME recorded this as faults.
- The leak was so strong that a drop in rail pressure is also possible. That would give an incorrect mixture in the first bank, and so faults for the probe signals of that bank.
I cannot exclude that debris in the fuel played a role. In any case, the problem lies in the fuel mixture and the injectors.
The dealer’s diagnosis
For some time the car behaved normally, but a week later the second failure occurred. This time the owner decided to take the car to the BMW dealer.
What do you think the dealer decided? Their conclusion was that the valve seals had to be replaced, at a cost of EUR 3700. As their evidence, they pointed to oily spark plugs in cylinders 1 and 5.
My only question is whether BMW really hires people without education or experience for its dealerships. Is there nobody, no experienced colleague, who could help such a person? Is EUR 3700 so small a sum that it can be wasted so easily?
If the dealership is reading this, please explain the following.
First, cylinder 1 was working and was never switched off. How could its spark plug be oily? I will explain: the oil ran onto the spark plug when it was unscrewed, because the valve cover gasket is leaking. It really is that simple, and the same applies to the spark plug of cylinder 5.
Second, suppose the oil really was running down the valves onto the spark plug. I must disappoint you: replacing the valve seals would not help. There would have to be a huge amount of play in the valve guide, and the cylinder head would need repair. I cannot imagine how colossal the damage would have to be for enough oil to leak along the valves to “drown” a spark plug.
Third, how do you explain that an engine with such a valve seal defect runs perfectly, suddenly “breaks” for 10 seconds, and then runs perfectly again? Will you really claim that such mechanical defects are sporadic?
Fourth, are you confusing the fault entries for injectors, switched-off cylinders and Lambda probe readings? And note that it is precisely the second bank. Cylinders 1 to 3 are fine, with no misfires detected. Yet the valve seals supposedly failed in exactly cylinders 4 to 6, damaged the injector and the Lambda probes, and then everything recovered a moment later?
I think any technically minded person can see that the valve seal theory is entirely inadequate. But how could this happen at a BMW dealership? How?