Bear with me for a moment. The inrush current to a coil can be many times greater than the nominal load for the coil. Coils (such as starter windings and solenoid windings) have almost no resistance until their magnetic field stabilizes and the coil's inductance begins to function. This physical phenomenon enables a huge discharge across the starter posts inside the solenoid. This discharge can be anywhere from 10 -50 times more than the nominal amperage rating for the coil (which, let's say, is approx 150 amps). In our vans, this discharge is only limited by the CCA of the battery.
This is completely
incorrect. Solenoids in DC circuits have NO inrush current. They are inductors whose current builds from 0A to a maximum value determined by the DC resistance of the inductor. See:
http://www.electronics-tutorials.ws/inductor/lr-circuits.html
You are confusing the starting characteristics of a motor with those of a solenoid/coil/inductor. A motor consists of windings which are inductive, but their resistance is low so the current builds very quickly to a large value (the inrush current) which is determined by the total resistance in the circuit = winding resistance + resistance of the cabling (Battery to starter + grounding cables) + internal resistance of the battery.
Once the motor starts to rotate it produces what is called a back EMF which opposes the flow of current, eventually reducing the current to the 'normal' running value when the motor is operating at its designed speed & load. It is then the amount of LOAD (the power required to rotate the diesel engine) that is the main determinant of the amount of 'normal' running current.
A 950CCA battery will produce more inrush current and a larger flash during the inrush moment than say, a 750CCA battery. And given the small physical size of the internal contacts within these solenoids, 200 more amps surging during inrush may explain why I have had four solenoid failures.
The CCA capacity of the batteries is, in your example, irrelevant. The current delivered by the battery depends on the LOAD, not its capacity to provide current, with the proviso that the load does not require a current larger than the battery is able to provide. I would expect the current draw of the starter motor to be less than 500A max (no data, willing to be corrected).
In short, while the 950CCA battery
can provide 200A more current, it is highly unlikely to do that in the case of a starter motor, unless the motor is jammed, or otherwise malfunctioning.
Finally, in the case of inductive DC loads, such as a starter motor & a solenoid, the worst arcing will occur when the contacts
open, not when they close. The inductor (e.g. motor winding) stores energy in its magnetic field & there is mechanical energy in the rotating mass. This energy tends to keep the current flowing until the energy is dissipated, which causes an arc at the switch contact that is trying to stop the current flow.
If you care to look into the properties of inrush current, here's a good article:
https://en.wikipedia.org/wiki/Inrush_current
My hunch, after four starter failures and dissecting one solenoid, is that these Bosch solenoids are simply inadequate.
I will post some pictures of solenoids #3 and #4 when I dissect them.
Thanks!
Having typed all of the above, I'm sorry to say that I have no idea as to what is causing the premature failure of the starters/solenoids in your van. If this were a common problem, such as the Y-Cable & ground strap, I'm sure that others would have chimed in with a 'me too'. So, my conclusion is that there is something peculiar to your van, such as the 'tolerance stacking' mentioned in previous posts, that is causing the problem.
I hope that I have at least been able to correct some of your misconceptions regarding the electrical characteristics of the starter system.