I'm looking at a couple of stress reports that deal with the fatigue life of offshore Xmas tree flowlines. These flowlines typically have to be very flexible in order to survive an extended period offshore, and are thus usually hung from constant effort supports. The riser from the well passes through a 'conductor' which limits it's lateral movement caused by wave action before connection to the Xmas tree. Therefore a typical spec would be +/-25mm laterally [and sometimes +25mm vertically] caused by wave action and +100mm upwards caused by thermal growth of the riser itself. The wave action is very frequent [say 10 waves an minute, for 20 years = 105120000 cycles], but the thermal growth, although much larger, may only be 1 cycle evry month or so.
The reports I have seem have taken the stress range caused by the movement from full +ve to full -ve wave action in each direction from a static analysis and used that to compute the fatigue life [how that is done is of course is another matter].
For those of you who have never been offshore, imagine a several tonne block of steel with piping coming off it [the xmas tree], dancing around as if it was being shaken by a giant and you'll get the picture of what goes on.

If we assume that a full dynamic analysis is not nessecary, and that we cannot be sure that every wave WILL NOT cause full displacement, do we need to apply a DLF to the stress ranges caused by these lateral wave movements when they are considered in the fatigue calculation? Personally I think that it is not required for the main flowline, but may well be good practice on associated smaller lines, such as drains, vent valves and injection lines that may well get excited by the main flowline as it shakes during operation [dog shaking the tail, as JB would say]. The addition of a DLF to these branches may well be a case of conservatively increasing the stress ranges being considered and hence reduce the fatigue life of certain parts of the system, but is that nessecarily a bad thing?