DNV-RP-105 suggests that galloping has to be properly taken into account for piggy back solutions, however it doesn't give any clear guideline for doing this.

Is a CFD analysis/experimental test the only option?
Could anyone advise on how to consider it?

Heard it is more of an issue with risers than pipelines, but came across a report where it says onset of galloping could occur at reduced velocities as low as 4.

Any thoughts on this?

Flow across either pipe should not dynamicaly excite the piggybacked line.  More of a danger for the piggybacked line, since they are of smaller diameters and the impinging velocity may be higher, due to proximity to the carrier line.  The piggy may also be exposed to vortex shedding off that carrier line.

It isn't normally a problem for the pipeline portion, since non-shoaling waves have no orbital velocity at the mudline.  Risers are exposed to the highest orbital velocities.  Where shoaling waves are encountered, the pipelines are likely to be buried or rock, or mattress covered.

You would want the piggybacked pipe to not have a natural frequency between 0.7 and 1.4 times the Von Karmen vortex shedding frequencies at any orbital velocities encountered in the wave for that line, nor a frequency equal to any vortex shedding frequency from the carrier line either.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso