When you say "in the vicinity of the PSV" I trust you are refering to vicinity as in frequency rather than vicinity as in location. Modal response is not necessarily excited through a load applied at a location of high modal deflection. Your (load) participation factor is an indicator of the mpodal response to that particular load location.
Yes, it sounds like your system is responding as a rigid body so a static load should suffice. Apparently, your valve has a hard Y support under the branch connection? If so, your DLF (if any) can be calculated using the approach in B31.3 Appx. II. Or, you can find the CAESAR II mode of vibration that shows the PSV wagging back and forth and use that frequency to pick your DLF from the spectrum generated by CAESAR II.
That 90% rule falls apart when you have rigid response. Let's say you have a beam anchored at one end and you put an axial dynamic load at the other. You'll have to run out to an axial mode of vibration to get any mass participation. But the system is responding in a rigid fashion and a static load is appropriate; there is no dynamic amplification in structural response.
_________________________
Dave Diehl