Model A (111 stations, 888 ports) clusters in county seats where population density peaks. It puts heavy concentrations in Harris (Houston), Dallas, Travis (Austin), and Bexar (San Antonio) counties — urban cores that already have 424, 272, 706, and 222 existing stations respectively.
Model B (49 stations, 948 ports) spreads to where people actually need public charging: I-35 corridor exits, I-10/I-35 interchanges, and suburban hubs like Katy, McKinney, Round Rock, and New Braunfels. It sizes each station to real demand (4–20 ports) instead of the flat 8. It accounts for the 3,976 existing AFDC stations and the fact that 80% of suburban residents charge at home — those dense metro stations are serving people who don't need them.
The optimizer achieves 88.5% coverage with 49 stations — 12.4 percentage points better than 111 screening stations at 76.1%. It uses 56% fewer stations but sizes them larger, delivering 948 ports vs 888. Model A clusters in Harris, Dallas, Travis, and Bexar counties. Model B spreads to corridor counties along I-35 and I-10, putting stations where the 472,572 registered Texas EVs actually travel.
Model B: Greedy MCLP using 41,467 real AADT station locations as candidates, 3,976 AFDC stations as existing coverage, 25-mile coverage radius. 4,151 substations for grid constraint checking.