I have seen thermocouples hose water 50 feet, I have seen the shield braid on RG8/U cable hose wate farteher than that. BUt thermocouples are cheap and easy and adequate for many applications, that's why folks use them. I also had a situation on a house that we purchased where water ran out of the entry panel when it rained, because the cheap person who installed the entry cable did not use an extra foot for the drip loop.
Water Proof encloseures will fill with water brought in by cable and conduit problems, so the very best choice is to have a drain in the enclosure bottom. But be sure that water can't enter by the drain opening. Another method that will prevent water entry is to have a pressurized enclosure, with sufficient air pressure to prevent water entry. This has been used on communication cables for many years. It is an expensive option, but much cheaper than fixing the problems caused by moisture entry. Of course, you must use dry air or dry nitrogen for pressurizing. That might have worked on the tug boat problem.
It is important, if you want to keep water out of enclosures, to consider that the water entering the enclosure may not be in liquid form. Some seals are great for keeping out liquid water, but may not prevent humid air from entering the enclosure. When the temperature drops, the moisture in the air condenses and ends up as liquid water inside the enclosure. Solutions include pressurizing the enclosure with dry air or an inert gas, making the enclosure air tight (as well as liquid-tight), or adding a condensate drain if high humidity inside the enclosure can be tolerated.
Something to think about: If you make a sealed box with wires entering it and you daily heat this box to temperatures significantly above ambient, then at night cool it, you have made an efficient pump. If the box is vented, the vent needs to be in a location that draws in dry air. If it is not vented, the box will find a way to breathe expelling and replacing the air from somewhere.
I've had radar lines on a 90' sailboat mast dump water into a junction box that was sealed. Every fall when the mast was removed, the terminals in the box would be severely corroded and the wire needed to be stripped-back to find bare wire. The wire was sealed in the radar dome and sealed in the junciton box, however in winter the mast was stored horizontally with the dome removed and the wire exposed. Somehow water got into the cable and until it was replaced, caused corosion in the sealed box. Once you see this, you can find the same symptom in other areas. It makes for interesting trouble-shooting.
In my experience the scheduled cleaning of the service equipment was the biggest issue with corresion.I noticed that seals were not holdig with time and the quarterly cleaning let some moisture inside. This was the main cause for corrosion on power terminals and transformers, especially where the thermal cycling was evident.
Currently I am working on the characterization and performance evaluation of solar collectors, we are using thermocouples for temperatures measurment in a closed loop of solar collectors in order to read temperature elevation of a recirculation fluid. One of our headaches is water in the terminals of the sensors, it produces oxidation in the ferrite metal of the thermocouple and therefore lecture errors, if anyone can suggest an efective way to isolate thermocouple terminals from water it will be really appreciated.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.