I assume that the crane will be used mainly when the vessel is along side the wharf.
It may be possible to turn off selected equipment to allow enough capacity to start the crane. Any heating loads are a good starting place.
The KVA of the motor start will probably be at a very low power factor. The power factor of the base load will probably be at a much higher power factor. Due to the different phase angles of the starting current and the base load current the total current and the total KVA will be less than the sum of the parts. This is to your advantage but also complicates the calculations as to whether the set will start.
Take a look at the base loads and see what may be safely turned off for a few seconds to allow the motor to start.
I have found that the Cat programs are quite conservative. I ran the numbers of a few marginal installations that I had been stuck with that involved motor starting. (Residential with large A/C units) All of the installations were working adequately but the Cat software said that all would fail.
The Cat software is good for initial sizing but when you are stuck with a large motor to start on an existing set, it is discouraging.
Based on direct experience starting motors on existing gen-sets with existing base loads, your installation is a little worse than marginal. There are a few factors such as the ratio of the motor size to the base load and the phase angle difference between motor starting and the base load that work in your favor. If the gen-set is prime rated rather than standby rated that is a big factor in your favor.
At the end of the day, the installation is still marginal. I have found that load curtailing is the most dependable and cheapest solution. This may be manual with suitable instructions, semi automatic or fully automatic.
In residential applications I have had good luck with semi-automatic schemes. The selected loads are tripped off automatically and manually restarted.
This is a case where you must make several assumptions in order to calculate the possibility of starting the motor. Then you go out in the field and see which assumptions were erroneous if the motor doesn't start.
You may consider a trip to the vessel and some testing to determine what conditions are.
I don't think that a generator will be happy with either a VFD or a soft start under marginal conditions. Both of these devices have input current waveforms that are challenging to the AVRs of generators. We are back to assumptions as to whether there is sufficient reserve capacity for these devices to function properly.
Bill
--------------------
"Why not the best?"
Jimmy Carter