One of the issues with the first 3 generations of nuclear power plants was that they were mostly designed to use the cooling water once and much of it evaporated during that use.
SMR can be very different. HTGR can be designed to use little or no cooling water if it is designed around a Brayton cycle. This potentially means that cooling water is not an issue at all with something like a pebble bed reactor. [however, this is not the current direction on the designs being proposed].
LWR will require cooling water. They can easily be built for once through, but as seen in newer fossil plants building a circular donut shaped cooling pond that takes water from one end and puts it back in the other end of the pond, can allow natural cooling and reduced losses from evaporation (a full cover for the pond helps).
One use of the residual heat is to use it to heat greenhouses. Depending on the crop and location, residual heat can be useful up to 9 months of the year. Another is for a district heating system, but premises would have to be with a reasonable distance to the SMR to make this work.
LWR can be designed to use up to 50% less water per MWH, if the designs are done correctly without having issues with overheating or reduction in efficiency. Add a cooling pond of sufficient size, and the plant can recycle most of the water needed.
At the present time, designs of Thorium SMR have little advantage over LWR for cooling needs. Same is true of MSR and FNR.
SMR I - Announcement of Small Modular Reactor Series
SMR II - Major types of reactors
SMR III -Â Why is 300 megawatts the dividing line for SMR?
SMR IV -Â Foundation independent designs &Â Power Conversion
NEXT: SMR VI -Â Nuclear Fuel