Have you included the pressure effects in your analyses? Including the Bourdon pressure effect and less significantly the stress stiffening due to pressure (options in Special Execution Parameters) can significantly affect your nozzle loading, your pipe being 10" and 14" will produce significant pressure loads (ask a fireman!). These pressure effects can significantly increase your nozzle loads especially the moments even in the "cold" case (W+P1), and where the piping system is not suitably restrained by guides etc. as seems to be in your case. Because the failure is occurring "within seconds of start-up" before the piping system has "warmed-up", pressure effects are the more likely cause of the overload.
Regarding friction affects at supports, they MUST be included in any analyses especially where sensitive equipment is concerned, as mentioned by John Luf. I would carry out 2 sets of analyses, one including friction affects at supports and another set without friction. The real case will lie somewhere between these 2 extremes. If friction at supports causes nozzle overloads, place low friction pads at supports with the highest friction loads closer to the nozzle until the nozzle loads become acceptable.