It has already been pointed out that a given substance boils at different temperatures under different pressures; the boiling point being raised when the pressure is increased and lowered when it is decreased. In the case of water, which boils under atmospheric pressure at 212° Fahrenheit, an increase in pressure to 70 pounds gauge raises the boiling point to 316° Fahrenheit, and a reduction in pressure to 29.74 inches vacuum lowers it to 32° Fahrenheit, or to its freezing point. From this, since the law is a general one applying to all known liquefiable gases, it follows that to produce low temperatures the pressure on the refrigerating medium must be reduced to such a point that the corresponding boiling point will be a sufficient number of degrees below the temperatures to be produced to bring about the heat transfer through the expansion coils or other cooling surfaces. If, for example, it is desired to cool a cold-storage compartment to 10° Fahrenheit, a back pressure of 24 pounds gauge will be found too high to allow ammonia to boil at this temperature. At 23.64 pounds pressure it will boil at exactly 10° Fahrenheit, but since this is the temperature of the surrounding air, there is no difference in temperature to bring about a heat flow and the boiling will not continue. When the pressure is reduced to 19.46 pounds gauge, the ammonia will boil at 5° Fahrenheit, and at this temperature there will be sufficient inflow of heat from the 10° surrounding air to cause quite appreciable refrigeration. A further reduction to 15.67 pounds gauge lowers the temperature of the boiling ammonia to 0° Fahrenheit and the increase in temperature difference from 5° to 10° Fahrenheit will effect a rate of heat transfer just twice as great per square foot of pipe surface as was possible with half the difference. A still further reduction to 9.1 pounds gauge pressure will allow the ammonia to boil at — 10° Fahrenheit, at which temperature the heat flow from the 10° room will be twice as great as it was at 15.67 pounds and four times as great as it was at 19.46 pounds. In order to produce the same amount of cooling effect at 19.46 pounds pressure as was obtained at 9.1 pounds pressure, just four times as much pipe surface would have to be employed, and in order to do as much as at 15.67 pounds, just twice the surface would be required. If, instead of direct expansion, brine circulation is employed, it will be evident that for the same rate of heat flow, a lower pressure and temperature will be required in the latter case. Assuming, for example, that the rate of heat transmission per square foot per degree difference in temperature be the same between the ammonia and brine and between the brine and air as it is between the ammonia and air (an assumption which is not wholly accurate, but which will simplify the example), the heat transmission between the ammonia at 9.1 pounds pressure and the air at 10° Fahrenheit would be only half as great in the case of the brine system as in the case of direct expansion. This because of the fact that 10° difference in temperature must be allowed to cause a heat flow from the air to the brine and another 10° for the flow from the brine to the ammonia. On this basis, in order to produce the same heat flow, 9.1 pounds back pressure would have to be carried in the case of brine circulation against 15.67 pounds in the case of direct expansion. The relationships are expressed in tabular form in Table IX.
<Callout type=