There is another factor related to cooling water we need to consider. The river/body of water that the cooling water is released back into has to have enough volume to absorb the extra heat without dramatically increasing its temperature. Some power plants in the southeast are closed, not because there isn't enough water for cooling, but because the river ecosystem it's returning to would be completely killed off by 70/80/90/whatever degree water.

When droughts reduce the volumes in bodies of water, they can't accept the cooling water and the plant shuts down. So minimum required water levels for plant operation are significantly higher than you would expect just from looking at plant usage/evaporation.

Disclaimer: I do not work in power engineering.

However the way I understand it, Black_Dog is correct: a power station will have an "inner loop" driving the steam turbines with ultrapure water in a closed cycle (any crud in the water is hard on turbine blades). You could use any number of working fluids, anhydrous ammonia for example, but plain water steam has perhaps the highest latent heat content of any fluid. Hydrocarbon liquids tend to be sort of indifferent -- not much energy is moved about in changing between states.

Now, you could certainly employ a power station's water purification plant for desalination, etc. but it would cost some energy, much as carbon sequestration does. As for the open loop, its purpose is just to dump waste heat into the environment at the lowest possible temperature; thermodynamic efficiency and all that.

| _{The problem will solve itself.}
|  _{But not in a nice way.}

Quick answer: yes, it's possible, but very expensive in capital and loss of efficiency.

In power engineering there are 'topping' cycles, and 'bottoming' cycles.

Topping cycles are designed to use a fluid that can be handled effectively at temperature for which H2O offers serious difficulties. One fluid that I have heard of being used in this application is Mercury. It boils at higher temperature. The vapor maintains reasonably high density at very high temperatures. The condenser can be used to warm the H2O on its way to the main boiler. Condensing the Mercury is technically very easy. Keeping the piping absolutely leak free probably requires different gasket materials than H2O. Topping cycles have some importance in coal fired plants and less importance in nuclear plants because it is often hard to keep the reactor from melting at typical temperatures of topping.

Bottoming cycles are used in solar boiling water power stations. The fluid is one that boils at a lower temperature than H2O. Various fluids that have application in refrigeration systems also are used here. Boiling the fluid is done in the steam condenser. Recondensing the vapor is done in yet another condenser which itself requires either
cool air or cool water to carry away the reject heat. Bottoming cycles allow the use of a higher density vapor at low temperatures. This allow the use of physically smaller turbines.

I a physicist. I don't know much detail about these systems. There a lot of tricky ideas that have been tried to improve the efficiency of power plants. Part of the design process is figuring out how to build plants that have good survival under the stress of handling the nasty fluids and vapors. I suggest that great caution be exercised discussing this technology. A lot has been left out of THIS presentation. BEWARE!

Sorry, but when I first went to the csp site I found a neat little video that showed that an oil was used to carry the heat from the collectors to the power plant, at several hundred degrees. It doesn't appear to be there now, though there is one at Youtube on the subject. Pity since it showed off the project quite well, and I had thought I had referenced it.