The radiated power depends on the power introduced in antenna. For this, we define the anrenna gain rather than power. A patch antenna has typically a gain of 6-9 dBi. The rear power is less of 20 dB than the front power. The ground plan and the array abd the substrate can increase this value.
Do you mean antenna power is the capability of antenna to handle power? If yes, Micro strip patch can handle low power (of the order of mW or few watts. It depend upon break down property of substrate material used.
In HFSS the typical value of incident power to the antenna is 1W. this was normalized power. but i dont know in ADS my friend. i think for simulation all the software use normalized power.
"A Microstrip Antenna" is to vague definition to give you an answer. The maximum power a Microstrip antenna can handle is depending on losses ( heat ) and that depends on the substrate, Height and copper thickness ( frequency and skin effect? ).
So, based on the above, the heat from the losses will set the limitations for you.
With a ceramic substrate, good heat resistant bonding of the copper layer, a low loss and a high impedance - it should survive way more than 70W. Just try to to keep the Q value down as well to avoid excessive voltages.
But don't go and push this kind of power for a space application as you can't cool the antenna.
But at the end of the day. If you need tp push a couple of hundred Watts, make a proper test rig and push it...
The factors explained by Dan Andersson are taken care when you see the used material breakdown capacity. What to keep Q is not in the control. Value of q will depend upon material loss tangent.