I am pleased that some of you found my comments colorful.
My points were that equipment designed to be heated by steam usually is less efficient when exposed to a steam/condensate mixture. Heat transfer is decreased in a partially-flooded device, and slugs of condensate moving at steam velocity can be destructive. It is important to separate the steam from the condensate, and condensate forms continuously because heat leaks through pipe walls, even when they are insulated.
For the sake of grabbing real numbers: 100 psi dry saturated steam carries an enthalpy of 1189 BTUs per pound. 338F. WIthout losing any temperature whatsoever, it gives up 880 BTUs and becomes saturated condensate with an enthalpy of 309 BTU/lb. THAT's a pretty efficient heat transfer medium. You can reclaim the condensate at (initially) 309 btu/lb, but removing one BTU from a pound of the condensate drops its temperature by one degree (round number). True, those ARE BTUs that have been paid for and they are worth recovering. MOST of the piping in a steam-cycle power plant is dedicted to recovering heat from condensate and pumping essentially saturated feedwater back into the boiler. It still takes a lot of BTUs to boil it, but it's better than starting with cold water.
Every steam trap vendor with which I have had contact offers surveys to inspect steam systems, identify problem areas and malfunctioning traps, and to increase efficiency/decrease heat loss of the system. The biggest obstacle to doing this is that the expenses come out of different budgets and accountability is indirect. If a supervisor spends $10,000 repairing/replacing faulty steam traps, that comes out of his maintenance budget. The resulting $20,000 decrease in fuel cost gets an attaboy for the supervisor at the steam plant. There is not always a recognition of the cause-effect relationship.