152 REFRIGERATION AND REFRIGERATOR INSULATION ON BOARD SHIP. 
thus, by restricting the flow of air over the ice, they reduce the meltage to a mini- 
mum. The average purchaser of one of these refrigerators is more interested in 
economy of ice than he is in the temperature to be secured in the box. He assumes 
that the temperatures are all right, but he knows each month whether or not the 
economy is right, when the iceman’s bill comes in. 
Heat enters a refrigerated compartment through five avenues :—First, through 
the walls, 7. e., through the insulation; second, in warm goods; third, by the inter- 
change of air through the opening of doors and by air leaks through defective insu- 
lation or through defective doors, cold air being heavier and immediately flowing 
out when doors are opened; fourth, from lights or from the heat of the bodies of 
workers; and fifth, from any change of state occurring in the goods, such as freez- 
ing, fermenting, etc. 
In large rooms these various losses of heat should be analyzed separately. In 
small refrigerators, such as those for hotels, kitchens, private homes, the average 
galley refrigerator and so on, a rough rule that gives quite as accurate result as a 
more elaborate analysis allows for a certain number of British thermal units per 
cubic foot of refrigerator per 24 hours; for instance, 1,300 British thermal units 
per cubic foot for an average pantry refrigerator. If the box is located in a kitchen, 
as in a hotel or restaurant, it is safer to allow about 600 British thermal units; and 
if the kitchen is especially warm, as many galley kitchens are, 900 British thermal 
units are safer, especially if the boxes are small. 
For butchers’ coolers or large boxes in hotels, 7. e., large storage boxes from 6 
feet by 8 feet to, say 10 feet by 15 feet, a safe allowance is, approximately, 200 to 
250 British thermal units per cubic foot for 24 hours. In climates varying appreci- 
ably from the temperate zone, corresponding allowances must, of course, be made. 
In general, the size of machine for a given refrigerator has to be decided upon 
before the refrigerator is put into service. On the other hand, where an existing 
ice-cooled refrigerator is to be cooled mechanically, one check on the size of the ma- 
chine required is given by the amount of ice which has been used. This check is, 
however, more likely than any other to lead to wrong conclusions unless the fig- 
ures are properly analyzed. The following is a suggested scheme of analysis :— 
First, determine the average daily ice consumption for the summer months— 
say, July or August—from the actual weighing of the ice, or from the ice bill. 
Second, determine as accurately as possible the average temperature that was 
secured by ice in the refrigerator. This will usually be from 55 to 60° and, 
although it is commonly stated as anywhere from 40 to 45°, these lower figures are 
so seldom obtained day in and day out as to practically warrant the statement that 
they are never reached. It is only with a full ice chamber and with the box closed 
for long periods that such low temperatures can be secured, and even under these 
conditions the average temperature will still run well above these figures. 
Third, calculate the heat inflow through the insulation with the temperature 
in the box as actually observed or as assumed, and with the average summer tem- 
perature prevailing outside. 
