Start with the vehicles, not the cabinet slots
The battery pool must support the actual vehicle operation. Start with vehicle count, route length, energy consumption, shift pattern, charging opportunity and the required service level. The installed batteries on vehicles are part of the total fleet pool, but they are not immediately available for a swap.
Map a normal day in time blocks. Record when batteries leave, return and need to be ready again. The busiest hour is usually more important than the daily average because it determines queue risk and the number of charged packs that must be on hand at once.
Estimate daily battery energy demand
A practical first estimate multiplies each vehicle’s daily distance by expected energy use, then compares the result with usable pack energy. This produces an initial count of pack-equivalents consumed each day.
Treat the result as a starting point, not a final order quantity. Real fleets have different routes, weather, rider behavior, load weights and battery conditions. The operating schedule determines how often a returned pack can complete a charge cycle before it is needed again.
Add the charging pool and reserve
Each operating vehicle needs an installed battery. In addition, the station needs packs that are charging, packs ready for exchange and a reserve for uneven demand, maintenance and unavailable assets. The correct reserve is not a fixed percentage for every network.
A depot with predictable return times can operate with a different reserve than a public, distributed network. Use station-level data rather than relying on fleet-wide averages, especially when one location handles a concentrated peak.
Check charge time against swap demand
Pack quantity cannot compensate indefinitely for insufficient charging power. For every station, calculate how many returned packs can reach the required state of charge during the available dwell time while staying within the site’s electrical input limit.
Charging strategy also matters. Smart scheduling may prioritize batteries needed for the next departure, but it must still manage pack temperature, state of health and local electricity tariffs. The station should be modelled as a queue of batteries, not a static collection of slots.
Plan for exceptions before launch
A reliable battery pool includes a response for damaged packs, unbalanced returns, delayed riders, seasonal demand and packs removed for inspection. Define where unavailable batteries are isolated and how the station communicates an exception to the operating team.
Battery identity and state records should follow each pack across vehicle, cabinet and service activity. Without that visibility, a fleet can own enough batteries on paper but still have too few serviceable packs at the required location.
Validate with a controlled pilot
Run a pilot long enough to capture busy days, different weather and the real behavior of drivers. Track swaps per hour, queue time, state of charge at return, charge completion time, ready-pack availability and service events.
Use the results to adjust the pack pool, cabinet location and charging rules before network expansion. A supplier can then propose the hardware and software configuration from observed demand rather than from a generic slot-count assumption.