Condition of 72 Inch Pipe After Ten Years’ Use
Steel Pipe Line in Service in Water Supply System of Brooklyn Borough, New York City, Examined After That Period of Use—Extent of Corrosion
WILLIAM W. BRUSH
Deputy Chief Engineer
Bureau of Water Supply, New York Department of Water Supply, Gas and Electricity
IN 1919, the T. A. Gillespie Co. under a contract dated November 6, 1908, laid a 72-inch steel pipe line from Valley Stream to Amityville, Long Island, a distance of 83,000 feet. This pipe was manufactured by the East Jersey Pipe Co. in 30 foot lengths, (except on curves, where the sections are shorter). The longitudinal joints were lock bar type, while the circumferential field joints are single riveted lap joints. The pipe is made up of 7/16 inch plates, and was dipped in a bath of mineral asphalt coating, known by the trade name of “Sarco.” The thickness of the coating was from 1/16th to l/32d of an inch. The pipe was designed for an internal working pressure of 125 pounds per square inch, it being expected that water would be pumped through this pipe at sufficient pressure to be delivered directly into the Brooklyn low service distribution system, or into the Ridgewood reservoir, which is the equalizing reservoir for this service. The pipe line from Valley Stream to Brooklyn, a distance of 43,000 feet, was laid in 1907, is 72 inches in diameter, 5/8th inches thick, with double riveted longitudinal and single riveted circular joints. Before the pipe line was completed it was decided that the original plan of pumping directly into the distribution system would not be economical due to the short time that would elapse before the Catskill supply was to be made available. Upon the introduction of the Catskill water the Brooklyn system was to be placed in reserve, and as new pumping stations were required to deliver the supply directly into the distribution system they would not operate long enough to warrant the added cost of construction.
The easterly end of the pipe line was placed in commission shortly after its completion, but was used to deliver only from one to five million gallons daily until the spring of 1917, when the Catskill supply was furnished. Since then the delivery has been about 100,000 gallons daily, to maintain the level of two ponds.
The condition of the interior and exterior of the pipe line was examined in 1914, and again in December, 1919, and it is believed that the story of the coating of the pipe and the condition of the metal after ten years’ use will be interesting.
After the pipe had been tested in the shop to a hydraulic pressure of 200 pounds per square inch, the specifications required that the pipe be then thoroughly cleaned and heated to 350° to 400° F. “after which it is to be dipped vertically in a bath of mineral asphalt pipe coating, equal to that manufactured by the Standard Asphalt and Rubber Company of Chicago. This coating must be durable, smooth, hard, tough, perfectly waterproof and strongly adhesive to the metal, and must show no tendency to flow when exposed to the sun in summer, or to become so brittle as to scale in winter. It must be free from blisters and bubbles. The bath must be heated in such a manner as to insure a constant and even temperature of 400 degrees Fahrenheit. The coating shall be not less than 1/32 inch and not more than 1/16 inch, thick, and must be a high resistant against electrolytic action. The material used for this coating must be shown by analysis to contain not less than 99.3 pure bitumen, consisting of not less than 70 per cent, petrolene (soluble in petroleum ether), and not more than 30 per cent, asphaltene (soluble in chloroform).”
The pipe was heated by an inside burner, the iron reaching a cherry red color, and was then swung vertically and lowered into a tank where the coating was maintained at the requisite temperature by superheated steam circulating in coils. Each pipe was usually dipped twice and allowed to drain for about five minutes, and then stood on end to dry. In the coating as applied none of the blisters subsequently noted were in evidence before the pipe was laid. The specifications provided for very careful handling of the pipe in shipment and laying, and if the coating was in any way damaged a field smear coating of the same material as the shop coat was to be applied cold and to be of approximately equal thickness. Two coats of this asphalt paint were to be applied to all field joints and to any abrasions of the coating which occurred during laying. The specifications were carefully carried out, and the field joints appeared at the time to be well protected by the asphalt paint as applied, but the subsequent corrosion of such joints proved otherwise.
Interior of Pipe
The first examination of the interior of the pipe was made on May 15, 1914, the pipe being entered at Massapequa at a point about two miles from the easterly end of the line. The second examination was made at the same point on December 15 and 16, 1919. Several flashlight pictures were tried but only one gave sufficiently clear results to be reproduced. The conditions found on one foot square typical sections are given in detail on the sketch, Fig. 1.
In general there was a fairly even distribution of tubercles and blisters over the entire interior surface of the pipe. The tubercles averaged about 1/4 inch in diameter and were soft. A relatively few large tubercles about two or three to a length were as large as 1 1/2 inches in diameter and 1/2 inch high. The metal under these large tubercles was eaten away to a maximum depth estimated to be slightly less than 1/16th inch. The coating appeared to have separated from the pipe at many small blisters, which to the eye apparently covered a much larger area than the 5 per cent, shown by careful measurements, and by counting the blisters. It is most important that in an examination of pipe surfaces the conditions be determined by actual measurement or else the results may be very misleading. The coating over each blister was brittle, and when a knife edge was drawn through the blister the coating frequently flaked off and in general showed clean metal underneath the blister, the space being filled with clean water. Incipient rusting on the edge of many of the small blisters was noted.
A-C-D represents areas 12″xl2″ laid off and marked for future reference, a count was made of the number of blisters or tubercles within each of these areas. B shows a 12″xl2″ square laid off on the coating, but not marked, the number of blisters or tubercles was estimated to be the same as at A.
Say 4 per cent, damaged; to this 4 per cent should be added areas of lock bars and riveted seams, which are rusted.
Total damaged surface……………. 4%
Including lock bars and riveted joints.. 3%
For a foot or so, each side of the lock bar joint, there were frequent large blisters, usually elliptical in shape, with the long axis one to two inches in length. The condition of the metal under the large blisters was about the same as under the small blisters. The entire coating had lost the greater part of its elasticity but was still functioning well in preventing corrosion. On and immediately adjoining the lock bars there were more tubercles and rust than on the remainder of the pipe. At all the circular joints, which were those made in the field and coated with the asphalt paint, there was heavy rust and tubercles, the field coating having been entirely obliterated. See Fig. 2. The relative corrosion of the metal when coated by dipping hot, as compared with a cold coating, was also afforded by the steel castings used to connect up the gates. These castings, which were cold painted, were entirely covered with tubercles averaging one inch in diameter and 1/2 inch high, and there was no evidence of the coating which under the specifications for these castings was to be “rust proof.” The metal showed pitting to a maximum depth of about 1/16 of an inch.
Exterior of Pipe
An examination was made of a section of the pipe about five feet square, from which the earth covering had been removed. This earth cover was of sand three feet in depth. The coating was found to be in excellent condition and showed substantially no deterioration. Whether the exterior coating would be found in such good condition in those sections where the ground water level stands above the bottom of the pipe is a question which cannot now be answered.
Comparison of 1914 with 1919 Conditions
There is no record available of exact measurements taken in 1914 and a comparison can only be made from recollection of the appearance of the surface of the pipe at that time. There was no very marked difference in the general appearance of the interior surface, except at the circular joints. The rusting of these joints in 1914 was not beyond the incipient stage. The number of rust tubercles appearing on the coated surface had perhaps doubled, and the large tubercles had increased in size and depth. The general rusting of the cast steel sections had occurred between 1914 and 1919.
The exterior portion of the pipe was in such good condition at the time of both examinations that no difference was detected.
Effect of Corrosion on Life of Pipe
The heavy rusting of the circular joints would indicate that here is where the pipe would first give way, and it would be many years in advance of the main body of the pipe. Where the coating had been destroyed and a large tubercle had formed the progress of corrosion would indicate a life of over seventy years before a pin hole would develop at the spot in question, as the rate of corrosion decreases as the depth of the rust tubercle increases. It would appear that the circular joints should be cleaned and painted to lengthen the life of the pipe. The exterior of the metal is so well protected that no deterioration need be feared from the outside, except possibly where the ground water is keeping the pipe surface continually wet.
It is evident that some different form of preservative coating than that used for the circular field joints must be applied to secure a reasonably long life for such coating. Just what such coating should be is a difficult question to answer. Sample coatings were applied by the Board of Water Supply to adjacent sections of 66inch steel pipe laid by it in Brooklyn in 1914. When opportunity is afforded to examine this pipe further light on the subject will be secured.
The recently presented report of the publicity and efficiency commission of Toledo shows that the division of water produced a revenue of $808,108.70, and paid $106,012.50 to redeem water works bonds falling due and to pay interest on the remaining bonds.