Group captivity
What this case shows
- Feasibility is a property of groups of sinks, not of single sinks.
- A flexible sink must not take local power a tight group needs.
- When restrictions cannot all be honoured, the sink with the fewest permitted alternatives is served first and the deficit falls on the sinks that had somewhere else to go.
Each snapshot below is one set of readings on this wiring, drawn as a flow chart — how to read it.
Hall tight pair
bat_c idle. {bat_a, bat_b} exactly exhaust east+west, so the 200 W home load must be served entirely from the grid.
| Reading | W | Of gross | Share | Made of |
|---|---|---|---|---|
| Gross power | 400 W | 1 | grid 200 W · east 100 W · west 100 W | |
| Grid import | 200 W | 1/2 | grid 200 W | |
| Grid export | 0 W | 0 | nothing exported | |
| PV production | 200 W | 1/2 | east 100 W · west 100 W | |
| Charging power | 200 W | 1/2 | bat_a 100 W · bat_b 100 W | |
| Discharging power | 0 W | 0 | no battery discharging | |
| Standby power | 0 W | 0 | no standby draw | |
| Self-consumption | 200 W | 1/2 | grid 200 W | |
| Metering imbalanceok | 0 W | 0 | readings reconcile exactly | |
| Home base load | 200 W | 1/2 | unmetered remainder of consumption |
Unsatisfiable overlap
bat_c now draws 100 W and is captive to east alone. Captive demand (300 W) exceeds local supply (200 W): someone must be deficited.
| Reading | W | Of gross | Share | Made of |
|---|---|---|---|---|
| Gross power | 400 W | 1 | grid 200 W · east 100 W · west 100 W | |
| Grid import | 200 W | 1/2 | grid 200 W | |
| Grid export | 0 W | 0 | nothing exported | |
| PV production | 200 W | 1/2 | east 100 W · west 100 W | |
| Charging power | 300 W | 3/4 | bat_a 100 W · bat_b 100 W · bat_c 100 W | |
| Discharging power | 0 W | 0 | no battery discharging | |
| Standby power | 0 W | 0 | no standby draw | |
| Self-consumption | 100 W | 1/4 | grid 100 W | |
| Metering imbalanceok | 0 W | 0 | readings reconcile exactly | |
| Home base load | 100 W | 1/4 | unmetered remainder of consumption |
Why this case exists
This is the case the max-flow solver exists for, and it is worth walking through slowly.
bat_a and bat_b may each charge from either PV system. Neither is
individually stuck — either PV system could cover either battery. But together
they draw exactly what the two PV systems make, so as a pair they need every local watt. Any
allocation that gives local power to somebody else leaves one of them short.
Asking "what must this sink take from this source?" one sink at a time cannot see that. It is Hall's condition, which quantifies over every subset of sinks, and no per-sink patch reproduces it. Watch the home base load in the first snapshot: it is unrestricted and could physically take solar, but it is served entirely from the grid because the pair have first claim.
The second snapshot then overloads the same wiring deliberately, and that is where the model runs out of settled answers.