Start with hover power and usable energy. Nominal pack energy is only the starting point — voltage sag, temperature, discharge rate, propulsion efficiency and payload all change how much flight time you actually get.
Nominal energy is not usable energy
The energy printed on a pack is a theoretical figure at a specified discharge. In flight, voltage sag, internal heating, temperature, cut-off voltage and BMS strategy all intervene, so the energy you can safely draw is usually below nominal. Older packs and high discharge rates narrow that gap further.
Endurance prediction should first subtract energy you cannot rely on. How much return and emergency-hover time the mission needs, and how much power remains at low temperature and low voltage, should be defined by the operator. Treating all nominal energy as flight time makes a handsome spreadsheet that cannot survive wind or pack ageing.
The same pack does not give the same endurance on a different airframe. All-up weight, prop efficiency, voltage platform and flight-control strategy all shift hover power. When discussing endurance, state capacity, health and mission reserve as explicit assumptions instead of quoting a bare number of minutes.

Hover power must cover the whole propulsion chain
Energy travels from pack through ESC, motor and prop into thrust, losing something at every stage. ESC switching loss, winding copper loss, core loss, bearing friction and aerodynamic loss all enter the thermal budget. At hover the prop must produce thrust equal to all-up weight plus payload, so the operating point is set by the aircraft.
With a poor motor-prop match the system may make up for low thrust with more current. Winding and wiring losses rise, pack voltage sags faster, and the aircraft hits its power limit sooner. Comparing no-load motor figures will never reveal this — you need thrust and efficiency curves at the target prop, voltage and load.
Beyond propulsion, avionics, video link, payload, gimbal and cooling all draw energy. Attitude correction in gusts adds margin, while climbs and acceleration raise power briefly. List these loads separately and sum them segment by segment instead of hiding everything under one average current.
Get an explainable figure with the segment method
Split a sortie into take-off, climb, hover, cruise, work, return and emergency hover. Each segment has its own power and duration, and hover and cruise must be evaluated with payload and wind. Add them up, compare against usable energy, and you can see which segment drains the pack and which assumption needs testing.
Power figures can be built from bench tests, flight logs and controlled flights. A bench controls load and environment; logs show real attitude correction and pack sag. Cross-check the two, recording prop, voltage, weight, temperature and wind so every run can explain its own deviations.
The final endurance figure should come with its conditions and a stated tolerance. Where return-to-home must be guaranteed, add wind, payload and pack-ageing scenarios to the model and check the reserve still holds. What you end up with is not a marketing number but an energy boundary usable for mission planning.