Just evaporation from the surface of the pooled water flooding the buildings, which will not be significant, even if it wasn't at a humid oceanside location.
This has to do with the strength of the hydrogen bond in water. The energy require to vaporize the water is the same as hundreds of degrees of merely heating the water, thats why the starting temp has little effect on water boiling rate, etc.
Still I stand by my numbers, 2 MW continuously can evaporate your 50 tons of water per day, to one sig fig.
Note that we're not arguing about much. Sunlight is like a KW per sq meter so we're arguing about the energy from 2000 or so sq meters of sunlit area, which isn't much compared to a giant nuclear plant facility.
It just doesn't strike me as much of an engineering challenge to evaporate 50 tons/day given that size of a facility and the decay heat and the solar insolation.
Keep in mind this water is all indoors, so unless you have some openings you are blowing air in, there will be no evaporation whatsoever.
Well, something that can't run forever, stops, and its been close enough to forever that you can assume we're evaporating away 50 tons/day. Where else could it be going? But is that a reasonable evap rate from an engineering perspective?
50 tons is about 50K kilograms. Takes about a KWH to boil away a kilo of water so we're looking at about 50K KWH or 50 MWH per day. Now there's about 25 hours in a day so thats about 2 MW continuous boiling away water, or at least on long term average.
A big reactor runs in the GW range but decay heat starts around 10% and slowly drops over time. Actually a couple years in, most should be gone although the long tail is pretty long. So it was dumping out tens of MW continuously and now maybe much less. So we're looking at a couple MW generated, vs a couple MW to make that much water vapor. Seems reasonable. So I'm willing to believe from an engineering estimate, aside from solar heating and whatever, that 50 tons of water simply evaporates away.
We can argue volumetric, if a ton is about 1K kilos and 1 L water is about 1 Kg, that would strongly imply a ton of water is about a 1K liters or about one cubic meter. I suppose we could have gotten to this same point by talking about mL vs cubic centimeters vs grams. Anyway you're claiming a cube of water a meter on a side, stacked 50 tall each day, for a couple years. I say that most of that has to have evaporated away otherwise you'd have 50 meters deep of water covering the entire site after a thousand or so days.