Dear Mr. Delaforce,

I will try to explain what is the story behind the R coefficient.

A building is not a rigid under a seismic event because- first- it is uneconomical to build it up as a rigid. Unfortunately a building is flexible enough to experiment G's accelerations much bigger than the maximum recorded with an instrument on ground i.e. the peak ground acceleration. For a building code, the building must be modeled as a cantilever with mass concentrated at stories. A vessel may be modeled similarly, with mass concentrated in GC of elements. A fine "mesh" would help.
And no doubts, a cantilever deflects under the seismic event.

In case the first period of vibration is "short" enough (and a first modal frequency of a couple of Hertz means it is), ASCE 7 asks to calculate a seismic coefficient as CS=SDs/(R/Ie). CS is applied to W- seismic weight, so CS can be seen as an equivalent lateral acceleration (G's fraction) for a global calculation of (Base) Seismic force. However CS is not seen as an uniform acceleration over the building height.
SDS is roughly 2.5 times the peak ground acceleration specific to a site, including specific soil condition influence and corresponding to a probability to have that seismic event.

For Ie=1 as importance and R=1- as elastic response, CS=SDs. For example if PGA=0.4G, it follows that CS=1. In other words, this means the lateral acceleration is the gravity acceleration. It is dramatic in design and complete unrealistic to be considered and for a tall building never considered.

That's why we accept some supplementary help by other practical aspects.
For a building structure, R>1 is an instrument for Cs reduction, possible for a number of reasons. As the structure begins to yield and deform in-elastically, the effective period of response of the structure lengthens which, for most structures, results in a reduction in strength demand.
Furthermore, the inelastic action results in a significant amount of energy dissipation in addition to other sources of damping. This is a combined effect we call ductility reduction.

The only remark I could have is that R=2.5 as ductility factor will bring back Cs to the level of the peak ground acceleration but we need to be sure R=2.5 can be obtained. The R factors for buildings are obtained by observations and experiments and published for buildings, there is no formula to calculate them.

For our case, I think the above written explanations would rise the question of the degree of plasticity we accept when calculating the seismic effect of a vertical vessel and what exactly R=2.5 means for such case.