This section is from the book "A Treatise On Architecture And Building Construction Vol4: Plumbing And Gas-Fitting, Heating And Ventilation, Painting And Decorating, Estimating And Calculating Quantities", by The Colliery Engineer Co. Also available from Amazon: A Treatise On Architecture And Building Construction.
147. Earthen drains should be carefully tested for leakage before the trenches are filled. The low end of the line of pipe should be plugged, and all branches should be stopped temporarily. The drain should then be filled with water, and allowed to stand full for a few hours. If the water settles down in the pipe, the leak should be found and stopped. A pressure of at least 1 pound to the square inch should be put upon every joint and pipe in the system.
148. All the drain, soil, waste, and vent pipes within a building should be thoroughly tested by water pressure, or the hydrostatic test, as it is called, before they are enclosed, and before the fixtures are attached. The pipes should be tightly plugged near the main trap or where they pass out of the building, and the end of every branch should be stopped water-tight. The system may then be filled with water, and the pipes should be allowed to stand full for several hours unless they are subject to frost, when they should be emptied before the water freezes. If the water sinks, then every pipe and joint should be inspected, and the leak found and remedied. A little more calking will usually suffice to stop a leaking joint in cast-iron pipes. However, if a cracked or split pipe or hub is found, the pipe should be removed, and a sound one should be put in its place. Patching or repairing should not be attempted; honest and durable work can be done only by replacing the damaged parts with new and sound pipes.
149. The test by water pressure is applied only to the iron stacks, branches, and drain pipes; but it is just as important that the fixture connections be made gas-tight, so a final test is applied to them when the fixtures are all connected up and the traps sealed. The pressure of such a test must be less than that required to force the trap seals. To find whetherthesystemis gas-tight, a smoke test should beapplied.
A modern smoke machine, which is commonly used for testing plumbing, is shown in Fig. 58. It is connected by a rubber hose a to some pipes, which are joined as shown, merely for the purpose of illustrating the principles of the test. The smoke machine is essentially composed of a double-action bellows b and a firebox, which is shown, with fuel in it, at c. A water-jacket surrounds the firebox, and a cover or drum d is placed over the firebox so as to be sealed by water in the jacket, as shown. By the construction of the machine, it will be seen that if the lever handle e is moved sidewise, to and fro, air will be forced into the firebox, through the pipe which is provided with a three-way lever-handle shut-off cock i. The fuel in c, being ignited at the bottom, burns upwards, and the air supplies the oxygen necessary for a slow and incomplete combustion of the mass. Smoke, con: sequently, is given off from the surface of the fuel (which is generally old, greasy cotton waste gathered as scrap from the pipe vise) and is forced down through the hose a into the drainage system.
Now let the illustration represent a system of house drains, f being the main intercepting trap, g the main house drain, h the fresh-air inlet, j the soil-pipe stack, k the vent-pipe stack, l a fixture trap, and m its back-vent connection; let us proceed to test the system with smoke. We leave the tops of k and j open; then we blow dense smoke. into the system. The smoke will fall down h, roll along g, and then rise up each of the vertical pipe lines and so push the air ahead of it out through the open ends on top. When the air is all out, and the smoke consequently blows freely from the open ends in dense, heavy clouds, the ends are sealed airtight by laying a 6 or 7 pound sheet-lead cap over each, as shown at o and n. Of course, these caps must be bedded in putty or clay, so that they will be air-tight. The instant the ends are closed, the system is full of smoke at atmospheric pressure, and it will be easily seen that, if the bellows continues to pump air into the firebox, the smoke will be compressed in the system and its pressure will increase.
As the pressure is increased in the system, however, so, also, is it increased in the firebox, and it will continue to increase until the upward pressure on the under side of the drum d overbalances its weight, when the drum will rise and float in the water. The pressure required to raise the drum is equal to that required to support a column of water about 1 inch high. In the figure, the drum is floating, a pressure of 1 inch water column is on the system, and the cock i is closed, to prevent the smoke from escaping down into the bellows. The effect of the pressure is also observed on the water in the traps l and f
If any leaks are present in the system, they can easily be detected by smoke flowing from them, and the drum will descend with a velocity which will vary with the extent of the leaks. If everything is air-tight, however, the pressure will remain constant in the system, and the drum, consequently, will continue to float.
If the drum floats for a period of about 5 or 10 minutes, the system is usually passed as gas-tight.
If the trap seals are not forced with any pressure less than that required to raise the drum, the traps are passed as being safely sealed.
150. The peppermint test may be applied in the same way by putting some of the oil of peppermint into the apparatus instead of the smoking material. A more convenient way is to pour from 3 to 5 ounces of the peppermint into the top of the vent stack, and follow it up with a half gallon or so of boiling water. The hot water makes the peppermint more volatile and helps to diffuse the odor rapidly throughout the pipe system. All the outlets should be closed and the air pressure should be put on as before.
The smoke test is recognized as the most satisfactory test. It has the advantage, when properly applied, of exposing numerous defects which cannot be observed by other tests.