elcb and borehole pump

If the pump is modern, with some built-in soft-starting (or running) electronics, it may have a filter with capacitance L-E & N-E as well as L-N, so a brief current can flow L-E when turned on.  I've seen this trip at least older RCDs from e.g. a washing-machine filter, just at the time of turning on or off the socket. However, like several other speculations before it, this doesn't very satisfyingly explain why the RCD would show no sign of trying to trip when it, instead of the pump control, turns the pump on.  The obvious ideal is to view the start-up residual current, but I assume you don't have some fancy clamp and plotting-device. Given that RCDs are easier to change than pumps down deep holes, the earlier reply about a brand recognised as being more resistant to brief leakage might be worth pursuing, in spite of the cost.  The idea of N-E faults (for the rest of the installation too, with normally-connected appliances, not just the pump) that was already mentioned sounds worth checking with a simple insulation test, if the motor takes high starting current in spite of its low running current. A diagram of the setup (including pressure switch) and the tests you've already done would be helpful.

If the pump is modern, with some built-in soft-starting (or running) electronics, it may have a filter with capacitance L-E & N-E as well as L-N, so a brief current can flow L-E when turned on.  I've seen this trip at least older RCDs from e.g. a washing-machine filter, just at the time of turning on or off the socket. However, like several other speculations before it, this doesn't very satisfyingly explain why the RCD would show no sign of trying to trip when it, instead of the pump control, turns the pump on.  The obvious ideal is to view the start-up residual current, but I assume you don't have some fancy clamp and plotting-device. Given that RCDs are easier to change than pumps down deep holes, the earlier reply about a brand recognised as being more resistant to brief leakage might be worth pursuing, in spite of the cost.  The idea of N-E faults (for the rest of the installation too, with normally-connected appliances, not just the pump) that was already mentioned sounds worth checking with a simple insulation test, if the motor takes high starting current in spite of its low running current. A diagram of the setup (including pressure switch) and the tests you've already done would be helpful.