Dear Sean,
I’m not sure, but do you get the same problem if you initialize the pitch angle more appropriately for the wind speed, as we typically recommend: Initial conditions for tower and rotor? At 24 m/s, the mean pitch angle for the NREL 5-MW turbine should be around 22 deg; at 22 m/s, the mean pitch should be around 20 deg.
If you believe your wind file is of sufficient size to encompass the rotor throughout the simulation when it runs properly, than the error you are getting regarding the exhausted FF wind array is likely related to numerical problems associated with the large jumps in pitch angle, likely preceded by a warning about a “small angle approximation violation”.
Best regards,