Planning Urban Air Mobility: Vertiport Throughput, Fleet Sizing, and Charging Demand

The queuing and energy math behind planning an eVTOL air-taxi network — how cities size vertiport pad counts, fleet turnaround, and fast-charging infrastructure against projected passenger demand.

▶ Open the simulation

A new transport mode, an old planning problem

Electric vertical takeoff and landing (eVTOL) aircraft — battery-electric air taxis carrying roughly 2-6 passengers over 15-25 minute flights — are moving from prototype to early commercial service in several cities. The planning challenge they pose is structurally familiar to transport engineers: given a projected passenger demand, how many vehicles, how much ground infrastructure, and how much energy does the network need, and where does it bottleneck first? What's new is that all three questions (fleet, vertiports, charging) are tightly coupled in a way that bus or rail networks aren't, because each aircraft simultaneously needs a pad to take off from, a charge to fly, and a passenger to carry.

From passenger demand to flight count

The starting calculation is straightforward: daily passenger demand divided by vehicle capacity gives daily flights needed. At 4,200 daily passengers and a 4-seat vehicle, that's 4,200 ÷ 4 = 1,050 flights per day. Spread across a realistic 14-hour daily operating window (early morning through evening, not 24 hours, given noise curfews and demand patterns), that averages 1,050 ÷ 14 = 75 flights per hour — but average hourly flights understate the real design requirement, because demand concentrates in morning and evening peaks the way any commuter transport mode does, so infrastructure needs to be sized against peak-hour flights, commonly 1.5-2x the average hourly rate, not the flat average.

Vertiport throughput: a queuing problem

A vertiport's capacity is a function of pads per site, sites in the network, and how long each pad is occupied per aircraft turnaround (loading, unloading, safety checks, and — critically — charging, unless swappable batteries are used). At a 12-minute service time per pad, one pad handles 60 ÷ 12 = 5 flights per hour. Across 4 pads per vertiport and 6 vertiports citywide, total network throughput is 5 × 4 × 6 = 120 flights per hour — comfortably above the roughly 75-150 flights/hour range implied by the demand calculation above (average to peak), which tells a planner the vertiport network as specified is adequately sized, though the margin shrinks fast if service time creeps up (say, from slower charging or more thorough safety checks) or if demand grows past the current projection. This is a classic queuing-theory result: total throughput is the product of parallel service channels (pads) and the rate each channel can process, so the two levers for adding capacity are more pads (capital-intensive, needs land) or faster turnaround (needs faster charging and streamlined ground operations).

The energy demand behind the flight schedule

Each flight consumes energy proportional to distance flown and the aircraft's energy intensity — at a 32 km average flight distance and 1.8 kWh per km, a single flight uses 32 × 1.8 = 57.6 kWh. Scaled to the 1,050 daily flights calculated earlier, that's 1,050 × 57.6 ≈ 60,480 kWh, or about 60.5 MWh of electricity consumed by the fleet per day. Translating that into charging infrastructure requires knowing charger throughput: at roughly 120 kWh delivered per fast-charge session (typical of high-power aviation-grade DC fast charging), the fleet needs on the order of 60,480 ÷ 120 ≈ 504 charge-sessions worth of capacity distributed across the day and the fleet — in practice, planners divide this by the number of charging windows available per aircraft per day (since a given aircraft charges between flights, not once daily) to arrive at a realistic simultaneous-charger count, commonly landing in the 10-15 fast chargers range for an 85-aircraft fleet flying this schedule, spread across the vertiport network rather than concentrated at one site.

Why grid integration and V2G matter for the economics

A fleet drawing 60 MWh/day at concentrated peak-charging windows creates a meaningfully spiky grid load if all charging happens during the same afternoon lull between flight peaks — which is why serious UAM network designs pair fast-charging infrastructure with time-of-use tariff optimization and, increasingly, vehicle-to-grid (V2G) capability that lets grounded aircraft batteries discharge back to the grid during demand peaks in exchange for reduced charging costs. This isn't just an environmental nicety — at city scale, unmanaged simultaneous fast-charging of dozens of aircraft can trigger local grid capacity upgrades that are far more expensive than smart charging software, making load management an economic necessity rather than an optional feature.

The parts of the system that queuing math doesn't capture

Vertiport throughput and energy calculations describe the steady-state capacity of a well-functioning network, but real UAM systems also need airspace deconfliction (unmanned traffic management, or UTM, systems coordinating multiple aircraft sharing corridors near a vertiport), weather contingency (eVTOLs have tighter wind and precipitation operating limits than commercial airliners), and maintenance-driven fleet redundancy — planners typically build in a reserve fleet fraction (commonly 10-15% of the active fleet) specifically to absorb scheduled maintenance without degrading the peak-hour flight schedule calculated above.

Frequently Asked Questions

How is vertiport capacity actually calculated?

It's a queuing-theory calculation: throughput per pad (60 divided by the service time in minutes) multiplied by the number of pads per vertiport, multiplied by the number of vertiports in the network. A 12-minute service time gives 5 flights/hour per pad, so 4 pads across 6 vertiports yields 120 flights/hour of total network capacity.

Why does peak-hour demand matter more than average demand for infrastructure sizing?

Passenger demand for air taxis, like any commuter transport mode, concentrates in morning and evening peaks rather than spreading evenly across operating hours. Infrastructure sized only to the daily average would be overwhelmed during peak periods, so planners typically size against peak-hour flights — commonly 1.5-2x the average hourly rate — not the flat daily average.

How much electricity does a city-scale eVTOL fleet actually consume?

It scales directly with flight count and per-kilometre energy intensity: at roughly 1.8 kWh per km and a 32 km average flight, each flight uses about 58 kWh. A fleet flying 1,050 flights a day consumes around 60 MWh daily — enough to require dozens of fast-charging sessions distributed across the vertiport network throughout the day.

Why is grid integration a bigger deal for eVTOL fleets than it sounds?

If dozens of aircraft try to fast-charge simultaneously during a lull between flight peaks, the resulting spike in electricity demand can be large enough to require local grid capacity upgrades — a far more expensive fix than software-based smart charging or time-of-use tariff optimization, which is why serious network designs build in load management and sometimes vehicle-to-grid capability from the start.

Does the throughput calculation account for maintenance and bad weather?

Not directly — the pad-throughput and energy calculations describe steady-state capacity assuming every aircraft is available and flying. Real network planning adds a reserve fleet fraction, commonly 10-15% of the active fleet, specifically to absorb scheduled maintenance and weather-related groundings without degrading the peak-hour schedule.