I have now had a more detailed look at the Elearn diagrams for a 2008 van.
The headlamp height adjustment motors do indeed share a supply (via fuse F13) with the LH dip beam, so they will only operate when the dip beam is on.
For this age of van, the Dip and Main beam headlamps are controlled by relays, but there is NO current monitoring by the BCM, so ignore my previous comment about faulty bulbs etc. This also means that although an abnormal drain from the adjustment motors could blow F13 and leave you with one dipped headlight, it wouldn't cause both to be automatically be shut down by the BCM.
The finger of suspicion is pointing to the headlamp switch (stalk). If this is intermittently giving the wrong resistance value, this could be interpreted by the BCM as a command to switch lights off. Maximum resistance corresponds to "lights off", so an open circuit would also give this condition. The "active low" drive to the dipped beam relay will then be lost, the relay will open and both dipped beams will lose supply.
I don't have data for the various resistor values in the stalk, but if you unplug its connector and measure with an ohmmeter between pins 4 and 7 of the stalk I would expect to see for example 3000 ohms for OFF, 2000 ohms for SIDE, 1000 ohms for DIP and zero ohms for MAIN. These values should be "solid" i.e. no wavering. If this test is OK, I suggest replacing the stalk connector and repeating the resistance test at the BCM connector (unplugged from the BCM) so that the two stalk-to-BCM wires are included