Mechanical engineers dealing with metals and alloys know that unless you're near the top end, 50 F doesn't mean much and a 50 F design margin can be added without really influencing cost significantly. With seal materials however, 50 F difference in continuous exposure temperature can be the difference between adequate service life and quite rapid failure. A 50 or 70 C difference in temperature is a totally different world! You need to use care to avoid tripping across a seal temperature limit arbitrarily.
If the process is truly out of control in temperature by a significant amount such that significant excursions long enough to bring mean seal temperatures in valves higher than their limits for extended periods are a real possibility, then you must be conservative. In that case, chances are your system as a whole will suffer from other problems too and could use some additional work. But if, as in many systems, temperature and pressure are tied to one another (by the vapour pressure of a liquid system etc.), or protected by controls, then tighter limits can be meaningfully set for the system's P/T envelope and better sealing material selections are possible.
Too many designers are sloppy or arbitrary with safety factor selection, and in many cases safety factors are "chained" or arbitrarily added to by subsequent members of the project team, each in an effort to cover their own backsides. This can easily result in bad selections, where under-performing or expensive high temperature materials are selected needlessly.
If you eliminate a class VI shut-off valve in favour of a class IV metal-seated valve merely in an effort to be "conservative" on operating or upset temperature limits, you're not really doing the project a favour!
This kind of sloppy safety factor selection has, in my opinion, resulted in some "grade inflation" of some seal materials' stated temperature service limits over time. Whereas Teflon-containing mateirals were once universally agreed to be of little use when continuously above 450 F (232 C), many are now making claims about extending reinforced Teflon valve seat limits to 500 F (260 C). I've asked for but have never been given the testing data which shows that the up-rating is based on anything. Rather, I think what has happened is that people are using Teflon-seated valves in 450 F services with 500 F design temperatures and getting away with it because they never actually operate above 450 F.
