I don't like the idea of water that may flash to steam inside a porous material. Also, any chance to keep one side of the pipe dry, is far cheaper than wetting it on purpose.

The problem is the temperature and pressure in the presence of chlorides.

I believe a traditional closed lloop system has some advantages in gathering investment because it is proven technology.

The thing I keep having to remind myself is that maximizing efficiency of the heat provided is not what drives the cost down.

The heat provided is not like the heat provided by a fuel plant or even a solar or wind farm. Each of those carry a much higher degree of cost in the accessing the power, either in fuel costs, or photocell costs, or blade and tower costs.

The energy density is so high and there is no fuel charge, so the dominant costs are plant construction and maintanence. In this respect, a design that sacrifices carnot efficiency for cheaper construction and lower maintentance, will be more efficient in KW/$.

So other than a closed loop system....

A design using the leidenfrost effect might be employed to keep the temperature of the pipe below 250C (where inconel or incolloy or other high nickel alloy would better resist cl- stress corrosion cracking). leidenfrost effect describes the inefficient heat transfer process of film boiling (when you check a pan to see if it is ready to make pancakes, if the drops of water evaporate quickly it is not hot enough, but if they dance and stay around for a minute, then it si hot enough, that is the leidenfrost effect).

One way to employ this effect, a portion of the steam taken post-work and sufficiently cooled and relatively free to much of the impurities, yeilding relatively clean water, is pressurized, and injected tangentially into the pipe,, so that it spins on the polished inner wall as it mov3es up with the steam. Some very abrasion resistant hard material could also form flow control rings at intervals (imagine a taurus shape, just a bit smaller than the ID of the pipe with a teardrop cross section) to slow the speed of the boundary layer steam without causing excessibve turbulence.

If you can keep that boundary relatively smooth, you should be able to keep the pipe cool enough and free of abrasion. But it would require a stiff pipe, and that would be a lot of structure.

Another way to approach the problem is to convert much of the heat and pressure to kinetic very early while still near the ocean floor and drive a group of Tesla turbines with the high volume high velocity water at the top. Either one, or a series of eductor like peices could be fitted on the pipe, such that instead of high pressure high temperature steam, you have high volume high velocity water at much cooler temperatures. This would require a much larger pipe, but it is much cheaper to build a very large low pressure low temperature system, than a high pressure high temperature one, and the technology is here already. Tesla turbines have no blades per se, so are rugged and accepting of very mixed effluent. This has the benefit of involving no new technology, everything involved allows for years of working experince to be relied upon.

Another approach would be to convey long conducting rods through a structure built just to contain the flow of the vent (not pressurized). in service rods conduct heat to steam generators. While the rods are conveyed out of service, they are scraped to be cleaned and to mine the sediment.