Performing a hydrostatics analysis on a multihull in Orca3D should be no different than doing it on a monohull. Make sure your hull(s) is/are placed in the correct position in Rhino (transversely). You can either model the entire vessel (both hulls for a catamaran, all three for a trimaran, etc.) or model half of the vessel and click the mirror checkbox in the hydrostatics dialog. Like hydrostatics for a monohull, if you are doing an analysis where the deck will become submerged (such as a righting arm calculation) you will need to model the deck as well. If the deck or cross structure will not become immersed, then you don't need to model it; the hulls will still be treated as if they are rigidly connected.

When doing hydrostatics for a multi-hull, certain values like the block (Cb) and prismatic (Cp) coefficients may not appear to make sense. This is because the waterline beam (BWL) will be for the full beam, not just for one of the hulls. If you want to know the Cb or Cp of one of the hulls, you should run hydrostatics on just the one hull, and that hull should be moved so that it's centerplane is at Y=0 (assuming Y is your transverse coordinate) as if it was a monohull.

As always, be sure that the outward normal direction is correct for all of your surfaces (the outward normal should point out of the hull into the water).

In the image below, the stability is computed up to an angle of 12 degrees for a catamaran. In this case, both hulls were modeled, and the option for showing the resultant flotation plane has been selected. As previously mentioned, it would have been fine to only model one of the hulls, move it transversely to the correct location, and then check the Mirror option in the hydrostatics dialog.

In this image, the same model has had the righting arm computed up to 70 degrees. Note that the deck edge on the port hull has become submerged, requiring a deck surface to be modeled, but the cross structure area is not required because it would not be immersed.