Model A under-sizes high-demand stations: 8 ports for Harris County, which has peak demand of 74,877 kW? And it wastes budget on 111 scattered stations. Model B concentrates into 49 demand-matched stations — most top sites get the maximum 20 ports because the optimizer finds high-demand locations first.
Model B delivers 948 total ports vs Model A's 888, with 56% fewer stations. But the top 10 stations all need 20 ports at 150 kW each = 3.0 MW per station. That reveals a constraint the flat model never had to face.
Station-by-Station Comparison
| County | Peak Demand (kW) | Model A Ports | Model B Ports | Model B Power (MW) |
|---|---|---|---|---|
| Harris (Houston) | 74,877 | 8 | 20 | 3.0 |
| Dallas | 66,452 | 8 | 20 | 3.0 |
| Tarrant (Fort Worth) | 46,550 | 8 | 20 | 3.0 |
| Bexar (San Antonio) | 28,255 | 8 | 20 | 3.0 |
| Travis (Austin) | 14,265 | 8 | 20 | 3.0 |
| McLennan (Waco) | 3,800 | 8 | 6 | 0.9 |
| Potter (Amarillo) | 2,100 | 8 | 4 | 0.6 |
| Midland | 1,500 | 8 | 4 | 0.6 |
20 ports at 150 kW each = 3.0 MW per top station. That's the equivalent of powering 2,400 homes — a significant grid connection that requires a dedicated substation tie-in. Model B total: 948 ports vs Model A's 888 ports, but concentrated in fewer, larger stations.
Sizing: ports = ceil(peak_demand / port_power / target_utilization). 150 kW per port. Real demand from Census + TxDOT + TX DMV data. 4,151 substations checked for grid capacity.