We did this problem in undergraduate chemistry and I am very convinced of
the theory.  The fact is, as posted earlier, that the hot water melts the
exesting ice and makes the resulting sheet continuous and there is no
layering effect everytime that the Zamboni passes over the ice.  Now, in
contrast to the physical explanation given by others, the evaporation
effect has very little to do with the quick freezing of the hot water.
Cold water, placed on top of the ice, only contacts the ice at the exact
water/ice interface.  Hot water, on the other hand, melts down into the ice
so that the water/ice interface is much greater than just on top of the
ice/under the water.  Because ice is less dense than water, that means that
the molecular latice of ice has areas in it which are not filled and in
fact are empty.  The hot water molecules have a high energy and can, one
melt the ice, but, also invade the emply spaces in the ice latice.  When a
hot water molecule invades this latice, it is surrounded on 5-6 sides by
ice molecules.  This hot water molecule quickly loses it's energy to the
neighboring ice molecules and if the energy is not too great, then the ice
stays ice and the hot water molecule becomes ice and expands into the ice
latice.  Thus further hot water molecules can invade this new latice.  This
process continues as the hot water molecules move down into the ice and
finally each molecule has lost enough energy to become ice.  The heat given
off of the hot water molecules and absorbed by the ice molecules will raise
the temperature of the ice.  The refrigeration unit then cools the ice back
down to playing temp. In short, the hot water molecules, being surrounded
by ice molecules on almost all sides, are cooled quickly, whereas, cold
water molecules would only be touching ice on one side and the heat
transfer from cold water molecule to ice is 5-6 times less than it is with
the hot water molecules.  Evaporation may have a small effect, but the
Zamboni puts down many, many gallons of water and they do not evaporate,
nor lose enough energy through evaportation to account for the rapid
freezing of water.  To back up this statement, try freeze drying water.
This is taking room temperature water or cold water and evaporating it
under vacuum untill it freezes.  This is a very slow process because the
evaporation does not remove energy from the water very quickly.  In effect
the hot water in the Zamboni is is a surface phenomina and I would love to
know who was brave enough to look at his fellow rink managers and say
"Boys, I'm gonna put hot water on the ice, 'cause it's gonna freeze
faster".
Flame on.
 
Jim Lawrence
B.S. Chemistry   1988 Bemidji State University
Ph.D. Biochemistry    1995 Purdue University