Degradation-aware sizing and scheduling of a solar-integrated two-wheeler battery-swapping station with heterogeneous pack health

Battery swapping can make electric two-wheelers economical for high-utilisation gig work in India, and rooftop photovoltaics can cut the cost of the energy swapped. Existing sizing models for solar-integrated swap stations, however, treat every battery pack as identical, although shared packs age at different rates and a weak pack fails a long-distance rider while still serving a commuter. This paper couples station sizing with per-pack health. A closed-loop 5-min simulation of a Bhubaneswar kiosk gives each lithium iron phosphate pack its own Arrhenius cycle and calendar fade, resistance growth and temperature, and serves 30 riders in three duty classes under calibrated synthetic weather. A degradation-aware swap scheduling policy hands the healthiest pack to high-need riders and the weakest adequate pack to low-need riders on a 95th-percentile need forecast, charges from solar unless a least-laxity test requires grid energy, and lets the sizing loop choose the retirement depth. Designs are selected under worst-year service limits using 4,610 five-year Monte Carlo runs in screening, verification and independent reporting stages. An identical-battery planning model understated equivalent annual cost by 15% and worst-year mid-trip swaps by about 3 times. Health-aware dispatch reduced worst-year mid-trip swaps by 14–25% at negligible direct cost, which made deep retirement feasible: the best health-aware design met the service limit with 2 fewer slots and 5 kWp less photovoltaics, at 3.3% lower cost than the best health-blind design. The benefit rose to 9.3% under a tighter limit and vanished under a loose one, while wear-levelling dispatch never helped. Network-level rebalancing and hardware validation are left for future work.