SECTION III, 1916 [93] (FRANS RSC: 
Production of Ring Ice or Hoar Frost in Pipes. 
By H. T. Barnes, D.Sc., F.R.S.C., Director of Physics, 
McGill University. 
(Read May 17th, 1916). 
An interesting form of ice formation has recently come under 
the writer’s notice which has on occasions given rise to some anxiety 
in places such as cold storage rooms where a dry pipe sprinkler system 
of fire protection has been installed. 
Owing to the low temperature at which the cold storage rooms 
are kept, no system of fire protection can be employed which neces- 
sitates the retaining of water during the period of inaction. Various 
non-freezing liquids have been used in the pipes, such as brine or 
calcium chloride, but the practice now appears to be general to install 
what is called the dry pipe system. The dry pipe system consists 
of replacing all the water of the sprinkler piping by compressed air 
at about 40 lbs. pressure. The water is held back outside the cool 
rooms by a hydrostatic valve which is balanced by the air pressure 
in the pipe. As soon as a sprinkler head is opened the air rushes 
out lowering the pressure until the valve trips, and the water rushes 
through the pipe to the sprinkler in action. For many months 
and sometimes years no water enters the dry pipes, and consequently 
no accumulation of water can take place at any point unless supplied 
to the system from without. Usually just after a pipe system has 
been installed a hydrostatic test is made to discover the position 
of leaky joints. Immediately after, however, all the water is drained 
out and the pipes left as dry as possible. It is fair to assume, there- 
fore, that any water remaining in elbows or crevices will become 
converted into solid ice whenever the pipes are below the freezing 
point. Air is pumped into the pipes at frequent intervals when the 
pressure drops below a given minimum. There is always a very 
slow leak even in the best constructed systems which has to be cor- 
rected by introducing more air. The intervals of pumping vary 
considerably depending of course on the tightness of the system. 
In a good installation about once a week is sufficient. In poor sys- 
tems pumping once a day is often necessary. 
It has been found on examination that ice appears inside the 
pipes and grows on the parts exposed to the coldest temperature. 
‘This ice forms as hoar frost and encircles the inside of the pipe grad- 
Sec. III, Sig. 2 
