For cleanliness, as well as for obtaining the best results, indirect stacks should be hung at one side of the register or flue receiving the warm air, and the cold-air duct should enter beneath the heater at the other side. A space of 10 inches, and preferably 12, should be allowed for the warm air above the stack. The top of the casing should pitch upward toward the warm-air outlet at least an inch in its length. A space of from 6 to 8 inches should be allowed for cold air below the stack.

As the amount of air warmed per square foot of heating surface is less than in the case of steam, we may make the flues somewhat smaller as compared with the size of heater. The following proportions may be used under usual conditions: 1 1/2 square inches per square foot of radiation for the first floor, and 1 1/4 square inches for the second floor, and 1 1/4 square inches for the cold-air duct.

Pipe Connections

In hot water indirect work it is not desirable to supply more than 80 to 100 square feet of radiation from a single connection. When the requirements call for larger stacks they should be divided into two or more groups according to the size.

The branches supplying the stacks should pitch upward from the boiler to a point directly over the stack, then drop and make connection with the heater at such a point as the special form in use requires. An air valve should be placed in the highest point of the pipe just before it drops to the heater. The return should be taken from the bottom of the stack and carried at a lower level back to the boiler or heater.

Conditions may make it necessary to bring back several separate returns to the heater, but it is better practice to use one large flow main and a single return of the same size, branching to the different stacks as necessary.

Pipe Sizes

As the difference in elevation between the stacks and the heater is necessarily small, the pipes should be of ample size to offset the slow velocity of flow through them. The following sizes for runs up to 100 feet will be found ample for ordinary conditions. Some engineers make a practice of using somewhat smaller pipes, but the larger sizes will in general be found more satisfactory.