Priority Shower Boards

[SUNRAY]

IIRC, when we've tried to design something using interlocked contactors it looked like you could get problems with how it behaved on restoration of power.

No problems at all if the circuit is DESIGNED properly.
That one in the picture was not, it was thrown together in a hurry as a kneejerk reaction to sump water getting into places it shouldn't. While we were working there it was easy to manually manage the pumps but out of hours was a different matter. As it happened it stayed in use for months, much longer than initially predicted couple of days.
All those bits were literally laying around, even the enamelled copper wire was stripped off a burnt out transformer and the 2 red wires go to an hours run meter [from a scrapped boiler system] so they were able to monitor on a daily basis how close to flooding they were.

 

[John D v2.0]

The race is caused by a small delay in the circuit. to break it you need to use the delay in your favour, you would need a DT relay that flips over to B when B is using >7kw, and a separate ST relay that connects B when A is using <3Kw (could have a small time delay). Then when you retore power, A gets power, uses 7kW, and the second relay won't have time to flip over.
Would that work? It sounds like it would to me.

 

[ban-all-sheds]

Again caveated by "IIRC"/"AFAIR", an explicit delay introduced into one side of the apparatus was seen as the answer.

Please note I've not been having a go at Sunray - I was just curious, as he seemed to have built the thing collectively designed here as a shower prioritising gizmo, to know whether he modelled or tested how it behaved after a power cut ended.

If JohnW2 is still here, he might remember more details of the (potential) problem we foresaw.

 

[SUNRAY]

Again caveated by "IIRC"/"AFAIR", an explicit delay introduced into one side of the apparatus was seen as the answer.

Please note I've not been having a go at Sunray - I was just curious, as he seemed to have built the thing collectively designed here as a shower prioritising gizmo, to know whether he modelled or tested how it behaved after a power cut ended.

If JohnW2 is still here, he might remember more details of the (potential) problem we foresaw.

I haven't gone into any details so far as I have achieved this in so many ways depending on what is there to start with.
For example: [I'll stay with 2 sump pumps as thats where my input started]. 2 totally self contained pumps will be controlled very differently to 2 pumps with float switches controlling contactors.
But in essence you are correct that introducing a delay in one device basically gives priority to the other device, the delay is tiny, usually in the order of 50mS. I reduce it to the point that the system fails then double the time. Both systems will receive power/start for less than the duration of the timer.
In the following the salient points are contactors C10A & C10B and T6 added in to existing circuit including current switches CT7 & CT8. In this Pump A will get priority in event of both pumps demanding at power up or the unlikely situation where both start a demand at the same time. and both will be powered for the time it takes CT7 and C10B to react.