Dear "anubis512",

Considering the main question is how the horizontal and vertical components will be combined, the answer may be as the User Guide says in Combinations description: SRSS method is typically used to combine seismic directional components. In addition B31Ea is more precise when recommend how the SRSS must be applied.

How to make the step from "response spectrum" to "static" is just other discussion. You may start constructing the spectra (horizontal and vertical) as per applicable seismic code or other reference (FEMAP is a good reference for vertical spectrum, just google for fema_p-1050-1.pdf and can find a lot in C23.1 Design Vertical Response Spectrum) and see roughly what are the accelerations corresponding with your first modal response. Or you can just take the "peak" of all spectra with the idea that is the worst that can be coincident. After that you need to imagine (or not!) that there is a mechanism of reducing the level of seismic effects due to ductility. Here is another pain hidden in the coefficient "Rp": it means that my piping developed plasticity (contrary to the Code intent of limiting occasional stress!) or rather the supports "helped" me developing such plastic response? And I need to consider Rp just coincident with "ductility level of earthquake" or I can consider it in any level of seismic I want?

Without an answer, let's suppose you succeeded having two accelerations as
"equivalent" static: horizontal and vertical. You may go ahead as B31Ea says about SRSS, why not?

Now, to the next level of discussion: are the vertical and horizontal seismic components related somehow? Well, starting with Newmark's works in 70's, vertical acceleration is set how FEMAP says: "Historically, the amplitude of vertical ground motion has been inferred to be two-thirds (2/3) the amplitude of the horizontal ground motion. 2/3 = 0.67 and here is the point where this coefficient appears.
I don't know why engineers were so terrified by SRSS and turn to 100-40-40 or 100-30-30 rules, but let's check for a set of just two horizontal and vertical which are proportional with 1 and 2/3: SRSS(1,2/3)=sqrt (1+4/9)=1.20, 1+30%*2/3=1.20, 1+40%*2/3=1.26. Not bad, but really necessary?

Returning now to "User guide". With U2 set as vertical, I cannot see the logic to include there 0.67U2 instead to take SRSS! Is 0.67 coming from discussion about vertical?- I hope no.

I hope the above is enough to convince you to go ahead with method 2, because in method 2 the intend is clear: to combine seismic directional components!