Welding a Cracked 48-Inch Pipe Underground
Unusual and Difficult Repair Work on New York Main—Break Mended Without Exposing Pipe—How the Work Was Done
THE following is the description of an unusual and difficult job in welding performed on a 48-inch main under twenty feet of earth, without uncovering the pipe:
As a part of the Croton system, the city of New York has eight 48-inch cast iron mains running south from what is known as the 135th Street gatehouse, located at the corner of Convent Avenue, Manhattan. (See Fig. 1). Where the mains leave the gatehouse there is more than twenty feet of earth and rock fill over the tap of the 48-inch pipe lines, and there have been several cracked pipe under this heavy fill since the mains were laid nearly forty years ago. The cracks have been caused by the external loads, as the head of water is only about thirty feet where the mains leave the gatehouse. The pipes are the usual bell and spigot type, and have a thickness of 1 1/4 inches, or equivalent to about Class B American Water Works Association standard. Previous to 1925 it has been customary to excavate from the surface and remove any broken pipe. The last repair of this character was accomplished in 1923 at a cost of about $8,000. The high cost of repair is due to the character and depth of the excavated material and location of work. A cracked pipe was repaired this year by electric welding in place at a cost of about ten per cent of the cost of removal and replacement of the broken pipe. It is believed this repair job is the first one of its kind ever undertaken on a cast iron main and is therefore of unusual interest.
Break in Main
On February 28, 1925, a large leak showed in cellars south of the 135th Street gatehouse, and by shutting down successively the 48-inch mains it was found to be on the next to the most westerly of the eight 48-inch mains. On March 5 the manhole cover on this line, located in the gatehouse, was removed, and an examination of the inside of the pipe showed that the pipe was split along the bottom, extending from the bell end for a distance of 5.5 feet, the broken end being 77 feet south of the manhole. A substantial flow of ground water was entering the pipe through the crack. ‘Phis ground water is probably largely leakage from the gatehouse, and twelve mains that leave the gatehouse. Four of these mains run west and then south and are old Croton aqueduct mains.
Adopted Method of Repair
Consideration had been given in 1923 to the repair by welding of the main that had cracked that year, but the crack was on the top, and no welding of this kind of an overhead crack had been previously done. When it was found the new crack was on the bottom the question of welding the pipe in place was taken up. It was believed that electric welding was the only type that probably would not generate enough heat to melt the lead out of the joint, and the Electric Welding Company of America, foot of Court street,
Brooklyn, through its chief engineer, William Schenstrom, advised that the repair by welding was considered practicable, and that his company would agree to make the repair for $880, and guarantee a successful job, no payment to be made if the crack was not satisfactorily closed. After considerable discussion and investigation of difficult welding work successfully executed by this company, an order for the work was issued under date of April 24. The main was not urgently needed and there was no necessity to save time on this job.
The ground water stood at about the center of the main, and before welding could be undertaken the water level had to he lowered below the bottom of the pipe to prevent inflow and remove the cold water from contact with the iron to be welded. 1 wo 2-inch holes were tapped in the bottom of the broken length of pipe and short suction lengths of 1-1/2 inch pipe placed in these holes, and carried about two teet below the bottom of the pipe. These short lengths were connected to 2-inch pipe which was laid on the bottom of the 48 inch pipe to the manhole, and connected to a centrifugal pump driven by an electric motor. After about sixteen hours pumping the amount of water that was to be removed was less than that which would feed the pump, running continuously, and it was found necessary to substitute a diaphragm pump, as the centrifugal pump would lose its charge and would not then automatically pick it up again. The seepage, which may have been largely leakage from other mains, represented about 20 gallons per minute. The diaphragm pump was operated continuously throughout the time that the welding work was in operation.
Welding Work to Repair the Crack
At the end of the crack (Fig. 2) a three-quarter inch hole was bored through the pipe to prevent the crack continuing further along the pipe. Then a “V” shaped groove was cut in the pipe following the line of the crack. This groove was two inches wide at the top, and came to a point at the outside of the pipe. Manganese bronze was welded into this “V” cut and brought up to the level of the inside of the pipe. This welding would not give sufficient strength, as the welding metal does not adhere to the cast iron as strongly as to steel. It was therefore necessary to place a reinforcing steel plate on the bottom over the crack. This plate is 7/8 inches thick, and was machined to the curve of the pipe. In the plate and near the edges there are fourteen one-inch holes through which the plate was welded to steel stud bolts tapped into the pipe. Threeeighths inch stud bolts were placed in the pipe along the edges of the plate and cut off about flush with the pipe. The plate was then welded along its edges to these bolts, and the cast iron between the stud bolts. The photograph designated as Fig. 3, shows this work as completed.
(Continued on page 865)
Welding a Cracked 48″ Pipe Underground
(Continued from page 838)
Making the Joint Watertight
As the bell was cracked, as the lead might have moved in the joint, and as the welding might melt the lead, it was necessary to make the joint tight from the inside. At the top of the pipe the joint was opened up about one-and-three-quarters inches, and at the bottom about one-quarter of an inch. The upper part of the joint was too wide to be satisfactorily caulked with lead wool. A circular steel wedge or dutchman was made, one-and-one-quarter inches wide at its widest part, and tapered down to one-quarter of an inch at the two ends. This wedge extended about three-quarters of the distance around the pipe joint. The bottom of the joint, including a couple of inches of the ends of the steel wedge, was covered by a three-eighths inch steel plate seven inches wide, which was welded to the pipe by use of stud bolts in a manner similar to the welding of the bottom plate. From the end of the weld lead wool was used to caulk each side of the wedge to make a watertight joint. Fig. 4 shows the lower part of the joint, and Fig. 5 the upper part of the joint.
Time of Completion and Results
The work was completed on June 9, and required one hundred and one man days of skilled labor. On June 10 the water was again turned on, and the main continued in service until June 17, when it was shut down and the interior examined. There was no evidence of inflow or seepage of ground water, and the job was found to be in excellent condition. Since then the main has been used continuously.
Mr. Schenstrom, the chief engineer of the welding company, is entitled to a great deal of credit for devising the plan of welding, and successfully carrying it through under very difficult working conditions, and credit is also specially due to John F. Mead, superintendent in charge of repairs to the distribution system, Manhattan, who was in charge of this work, and very satisfactorily aided in planning and executing it.
(Excerpts from paper read before the September meeting of the New York Section, American Water Works Association at Albany.)
In discussing Mr. Brush’s paper Superintendent William A. McCaffrey, of the Oswego, N. Y., water department, asked how the welders entered the underground pipe.
Mr. Brush explained that a manhole had been constructed in the gatehouse, allowing access to the pipe. There had consequently been no expense in getting into the pipe; the only cost had been in pumping out the water.
Water Commissioner George C. Andrews of Buffalo, N. Y., inquired if the repair was in the nature of a section broken out of the pipe, or simply a crack.
Mr. Brush answered that it was a crack in the pipe. The steel plates referred to in the paper, according to Mr. Schenstrom, chief engineer of the Electric Welding Company of America, which had done the work and to whom Mr. Brush frequently referred the questions asked by the members, were used with the object of strengthening the weld.
Asked as to whether the same repair could have been done at the top of the pipe, Mr. Schenstrom said yes, but that such a job directly overhead would be much more difficult. The steel plates would have to be relied upon to strengthen the job. Copper alloy was used in the welding.
Asked by Mr. Little whether the work could have been done without the use of the plates, Mr. Schenstrom said it could, but that in case of a defect in the cast iron or a shift in the pipe the plates acted as strengthening mediums. One reason for using the plates was that the pipe was 40 years old; another that the pressure from outside might shift the position of the pipe and rupture it or the weld under certain conditions.