As usual, the simple momentum theory is our friend in these cases:
$P=\sqrt{\frac{T^3}{2 \rho A}}$
Considering that in hover each rotor lifts ¼ of the total weight $W$ and that we have 4 rotors, the needed power becomes:
$P=4\sqrt{\frac{(¼W)^3}{2 \rho A}}$
where:
- $W=8\cdot9.81=78.5~\mathrm N$
- $\rho=1.125~\mathrm{kg}/\mathrm m^3$
- $A=\pi(22\cdot0.0254)^2=0.98~\mathrm m^2$
Substituting the values we get a total power of:
$P=234~\mathrm W$
According to the source, the electrical engines work at a voltage $V=50.4~\mathrm V$. The power of an electrical machine is $P= V\cdot I$, which translates into a current of:
$I=P/V=234/50.4=4.6~\mathrm A$
The battery has a capacity $C=14~\mathrm{Ah}$ and delivering that current it drains after a time of:
$C/I=14/4.6=3~\mathrm h$
This is obviously a maximum theoretical value that does not take into account many factors like:
- electrical engines are not 100% efficient;
- some electrical power is used by the onboard electronics;
- batteries are normally not made deplete completely;
- for manoeuvring, more thrust and therefore more power is needed than for hovering; but less power is needed in forward flight;
- simple momentum theory underestimates the needed power of some 15% in hover; and at the typical Reynolds number of drones the efficiency is worse; I'd say that the underprediction might be some 25%.
Factoring in all these points, the given endurance might be plausible.