Document zoNENwMepV0Xwm67Kka9oqMZ3

. H. 'flR!MG S-23 HOUSTON REGION T>2 2 TO: 322, 322-S, 321, 321-S, 12 Division Headquarters January 25, 1962 PDA # 56 SAN ANTONIO S ATTACK ON THE PROBLEM OF CORROSION Skyrocketing maintenance and replacement costs due to corrosion of metal pipe in many parts of the country continue to grow in magnitude. Transite is finding greater and greater acceptance in these areas when it is necessary for a pipe to meet the challenge of aggressive soil and water conditions. Here are the highlights of the accompanying San Antonio story with reference numbers corresponding to the sections marked in the article. 1. By 1953 annual maintenance costs due to corrosion of mains, services, and meters which comprized 95% of the system, had passed the $400,000. mark. J 2. A special division within the Water Board was established to study corrosion. Soil corrosion was established as the major problem. 3. Greatest number of main breaks occurred in soils of resistivity less than 700 ohms/cm^. 4. Cast Iron mains may have a useful life of 10 years in 600 ohm/cm^ soil. 5. Protection of cast iron pipe with sand encasement is at best a stop gap measure. - 6. Cast Iron will have to be protected against corrosion in the process of manufacture to be a satisfactory product in ,,hotn soils. 7. Between 1,250,000 and 1,500,000 linear feet of pipe in the corrosive area should be replaced within the next 10 years at a cost estimated at $10,000,000. 8. Asbestos-cement pipe for sizes 8 inch through 20 inches are normally used in the corrosive or "hot" areas. , This product stability in "hot" soils can be used as one of our greatest selling tools in these areas where corrosion is a problem. It is extremely important in our long range selling program to develop data on the performance of both competitive materials and Transits in each territory. Statistics on maintenance and replacement costs, main breaks per mile per year, etc., are valuable information not only in selling locally but in building performance data which can be used on a national scale. Your continued help in seeking out this information bo that it can be correlated and used by all, will contribute to increased sales of Transits. Please include this San Antonio story under the Competitive Materials section of your Pressure Pipe Sales Book. RWS/bgs R. W. SMITH San Antonio's Attack on the Problem of Corrosion by Bruce E. Sasse* Acting General Manager San Antonio City Water Board San Antonio, Texas he problem of underground cor fronting the City Water Board of San T rosion is a problem that confronts Antonio, and the Board's aproach to almost every water works manits solution. charged with the installation and maintenance of a water works system, and in particular the distribution sys tem. The degree of the problem, natur ally, varies from place to place and, therefore, may be only of slight con cern to a fortunate few readers but of great concern to the majority of us. This paper will deal with the serious and very real problem of corrosion con Past History The first water mains were laid in San Antonio in 1878. Some of these mains are still in use and thus, the present distribution system has mains varying in age from those laid over 80 years ago to those recently installed. Difficulties and losses due to corrosion were experienced and recognized as such 50 years ago by the management *Rad bafor* th Southwest Section, American Water Works Assn. Convention hold in Galves ton, Texas, October 1960. of the San Antonio water system. Of fice records and filed findings 'reveal that special methods were long ago used to protect small diameter mains and lines. Mains and services installed in those early days were either cast-iron or gal vanized steel and many of the small mains, particularly the galvanized steel, were embedded in concrete or coated with tar and wrapped with burlap. Nevertheless, no attempt was made to protect the larger steel and cast-iron pipe installed over the years. The rapid deterioration of galvanized service lines however, caused the Board to discontinue their use altogether in FIG. 1 -- Diagram of maintenance expenses from 1951 through 1959; also anticipated expense. SOUTHWEST WATER WORKS JOURNAL for APRIL, 1961 FIG. 2 -- Incidence and cause of main breaks in ferrous pipe which, until 1957, comprised approximately 95% of San Antonio's distribution system. 1935, and since then all service lines installed have been of copper. Scope of the Problem It wasn't until 1953 that the prob lem of underground corrosion was given top priority. This was the year that the present General Manager, R. A. Thompson, Jr.*, was appointed to his post. The annual maintenance cost of mains, services and meters had passed the $400,000 mark by this time and the skyrocketing trend of maintenance costs, based on past ex perience, was indeed frightening. Main and service line failures were increasing at an alarming rate and our records and experience indicated that our principal cause of trouble was underground corrosion. Figure 1 dia grams the maintenance expenses from 1951 through 1959 and the trend of such costs to 1970, as well as the an ticipated expense following proposed improvements. Figure 2 diagrams the incidence and cause of main breaks in ferrous pipe which, until 1957, com prised approximately 95% of our dis tribution system. Similar trouble was being experienced with service lines, particularly with those in corrosive areas, many thousands of which were galvanized pipe services installed prior to 1935. Study of the Problem This skyrocketing trend of mainte nance costs, together with the knowl edge that approximately 75% of the main and service line failures was due directly to corrosion, prompted the es tablishment of a special division wltFin the Water Board organization ~m 1953 for the primary purpose o? studying the underground corrosion problem In preceding years several *Bob Thompson resigned Dec. 31, 1960 experts had been consulted and their In the San Antonio area, howeve opinion taken into account, but these these are minor problems. The greate: opinions had been based on a knowl problem results from soil corrosion c edge of corrosion in general and its * galvanic corrosion caused by differeri prevention rather than on actual knowledge of the particular conditions in the city. The program of the di tial electrolytes. Prominent among th causes of this type are non-uniforr texture of the soil in contact wit] vision placed emphasis on gathering information on the physical and chem ical conditions of the sub-surface en vironment. This information had to be taken and interpreted in such a manner that it could serve as a basis for conclusions as to: a. Specific cause of corrosion. b. Methods for prevention or at least reduction in the corrosion damage. c. Methods for prediction of the most suitable type of installation. d. Methods of surveillance of the corrosive action. Much has been learned about the problem, but a great deal remains un known and the search for answers still goes on. The types of corrosive action affect ing water mains can generally be clas sified into one of the following cate gories: ^ a. Stray current corrosion. b. Galvanic corrosion caused by dis similar metal connections. c. Galvanic corrosion caused by dif ferential electrolytes. Stray current corrosion during the days of the street cars created severe problems. Today, the most frequent cause of stray current corrosion involve pipe, differential aeration, difference in moisture content around the pipe and the presence of soluble salts. The corrosion of buried pipes is ai electro-chemical process which can fcx illustrated by the action of a wet-cel battery. An elemental battery consist: of two pieces of different metals, con nected together by a wire and im mersed in an electrolyte. The difference of electrical potential between the two metals, causes a flow of current to take place through the circuit thus formed, and a deterior ation of one of the metals, called the anode, and the deposition of the par ticles on the other metal will occur. When pipe is buried in moist soil it may be pictured as containing a large number of these elemental bat teries in each of which there will be a continuous loss of metallic particles from the anodes being deposited on the cathodes. One of the most significant out growths of studies of corrosion by the San Antonio Water Board has been the confirmation of the general rela tionship between soil resistivity and corrosion. stray currents produced as a result of It is known that a potential differ cathodic protection. Galvanic corrosion ence causing flow of current may oc caused by dissimilar metals, results cur between two points on the same from service connections where brass pipe due to variations in the compo- and copper materials are employed- in . sition of the-pipe or the surrounding contact with the cast-iron, and in con-, soil, and the rate of corrosion will de nections of steel and, in particular, pend on this difference in potentia galvanized steel pipe to cast-iron. and the rate of flow of the current. SOUTHWEST WATER WORKS JOURNAL for APRIL, 1961 n----1----------------------1--------RELATION Of RESTSTMTY TO MAIN BREAKS tBSr-IMt f12 1 r 1 1 t / / *21 mm. MMtntrr FIGS. 3 and 4, left, show meter and augers used. FIG. 5, above, is diagram of relationship between main breaks and resistivity ranges. It follows that the current will de pend on the kind of metal, the chemi cal characteristics of the soil and the presence of moisture in the soil. The conclusions can therefore be drawn that soil resistivity should be a good yardstick to measure corrosivity. This is particularly true if the soil is alkaline. In acid soil the soil resis tivity may not necessarily reflect the corrosiveness. However, in San An tonio the native soils are almost all alkaline. Soils contain a large number of chemical elements, many of which are present as insoluble compounds which have little chemical influence on cor rosion. In general, the more corrosive soils contain considerable amounts of soluble salts. It is generally accepted that most soil corrosion of pipe is electrical in nature rather than the result of direct chemical reactions. The soluble salts in the soil affect the rate of corrosion because they control the resistance of the path of the corrosion current and the potentials of the corrosion cells. Using resistivity meters and soil au gers, resistivity measurements have been made throughout the area served by the Water Board with readings taken at approximately 2500-foot in tervals and at depths varying from ground surface to 4 feet. The resis tivity meter used by the Board is a Model 243-A Vibraground meter. The unit consists of a single probe rod which is inserted into the ground at the desired depth and the resistivity is read directly on the meter. Figures 3 and 4 show the meter and auger used. The units of resistivity used are ohms per square centimeter per centimeter, or for convenience years in a 1200 ohm-cm area, 45 years ohms/cm3. in a 4000 ohm-cm environment and For the sake of brevity the units of perhaps 75 years in a 6000 ohm-cm soil resistance will be referred to as soil. ohm-cm in this discussion. These read Based on judgment, an analysis of ings were plotted on a map and the costs and other factors, the Water resistivity contours sketched in. Board has set a soil resistivity of 700 The locations of main breaks were ohm-cm as a limiting value, below superimposed and it was found that which protective measures must be the greater part were within those taken. areas in which the soil resistivity was With the ever mounting cost of re less than 700 ohm-cm. placements and repairs due to corros Figure 5 shows the relationship be ion damage, there are grave doubts tween main breaks and resistivity that this value is high enough. ranges. Figure 6 shows that 88% of the It can be seen that the great ma main breaks in ferrous pipe occur in jority of breaks occurred within the soils having a 700 ohm-cm resistivity 300 to 600 ohm-cm range, and it or less. This includes those breaks re should be emphasized that all breaks sulting from soil shifts as well as those on ferrous pipes form all causes are directly attributable to corrosive causes. included. On the- basis of evidence gathered A map of San Antonio has been to date, the most corrosive soils in San prepared showing the areas of danger Antonio are the finely divided black ously low resistivity. These areas exist and yellow clays. Of them, the yellow within 77 square miles, or 48% of the clay with a resistivity of 200 to 300 city's 160 square miles. The total pop ohm-cm appears to be the most cor ulation living within this "hot" soil rosive. area is approximately 340,000 people. .The effect of moisture content and With the aid of the resistivity map its' seasonal or cyclic fluctuations be it is now possible to predict the degree tween wet and dry periods has been of corrosiveness in any particular area reviewed. within the city. Consequently, an in Most of the soil samples taken for stallation can be planned with the cer analysis and study have been found tainty that it will be the best possible to contain about 20-25% moisture at and most economical for the conditions normal pipe depth, .even during the at hand. dry summer periods, therefore, it is It should be understood that no area felt that variations of soil moisture at of the city is entirely free of corros different times will not have any se ion although in areas of high resisL rious affect on the value of the resis tivity, corrosion takes place at a lower tivity readings. rate. This reasoning is based on the fact )For example, indications are that a that research has shown that in the cast-iron main may have a usetul life range of 20% to 60% moisture con of 10 years in 600 ohm-cm soil, 25 tent, the resistivity is fairly constant. SOUTHWEST WATER WORKS JOURNAL for APRIL, 1961 Of interest, and indirectly related completely graphitized during the to the corrosion problem, is the fact ame period of years. that these corrosive clays are also highly cohesive and subject to extreme variations in volume with changes in moisture content. The resulting enor mous soil stresses make the structural strength of the pipe of paramount im portance. Over 60% of all breaks occur in the 30% of 4-inch and smaller pipe present in the distribution system. Figure 7 shows the incidence of main breaks per mile of main according to size grouping. With a knowledge of the causes of corrosion, the problem of protection in an economic manner arises. Field tests were initiated and evaluated to ascer tain the most practical solution to the problem. Although encasement of cast-iron pipe with sand appears at this time to be the best and most practical method at hand for the protection of cast-iron pipe installed in corrosive soils, it is at best a stop-gap measure. The danger of lumps of corrosive soil being accidently left against the barrel of the pipe during installation or later during the tapping of the pipe for services is ever present. Several test sections of cast-iron pipe coated or wrapped with insula ting materials were installed two years ago. These test sections were furnished by the manufacturer of the pipe and consisted of several joints of 6-inch pipe, factory-coated with a polyethe-' lene and sand coating. Test sections of cast-iron pipe in (^Several additional lengths of 6-inch cased in silica sand were installed in the most highly corrosive areas some five to seven years ago. In September, 1959, parts of these sections were un covered for inspection. They showed no evidence of corrosion pits where the sand is in contact with the pipe. pipe loosely wrapped with a 10 mil polyethelene wrap were also installed. Based on our own experience, we are of the opinion, that cast-iron pipe -will have to be protected against corrosion in the process of manufacture, in order to be a completely satisfactory product Unprotected cast-iron pipe installed for installation in aggressive or "hot" in similar areas has been found to be soils. Based upon studies and analyses over the past years of all main breaks, we have reached the conclusion that between 1,250,000 and 1,500,000 linear feet of pipe in the corrosive areas should be replaced within the next decade. The cost ot replacing these badly corroded mains and appur^ tenances is estimated at approximate^ $10,000,000. As previously stated, the story of increased maintenance costs due to corrosion is frightening; the number of main breaks and service line failures due to soil corrosion has climbed alarmingly over the past five years. We are keeping mains repaired and operating by sheer frantic effort and at great cost to the system and to the customer. It is now realized that only a costly replacement program promptly initi ated and pursued year after year will change the skyrocketing trend of main tenance costs. Nevertheless, this is the price that must be paid if the system is to survive both economically and physically. Figure 1 showed how this trend can be changed. The failure to face up to the seriousness of the corrosion prob lem years earlier than was the case is FIG. 6, left, shows that 88% of main breaks in ferrous pipe occur in soils having 700 ohm-cm resistivity or less. FIG. 7, below, indicates the incidence of mtain breaks per mile of main according to size grouping. 100 100 300 400 NUMBER OF BREAKS DISTRIBUTION OF BREAKS BY RESISTIVITY RANGE (ALL FERROUS PIPE) SOUTHWEST WATER WORKS JOURNAL for APRIL, 1961 an unfortunate one, but with die economy of the pipe plus the benefits knowledge now at hand and given of its protection from corrosion. To funds for carrying out the recommend reduce the opportunity for either the ed program for the replacement of cor steel cylinder or .the reinforcing steel roded mains, the picture will have wires from coming into contact with changed for the better by 1970. the soil, taps on this type of pipe are Attacking the Problem To protect our new installations against the ravages of corrosion, pro tective measures have been included in our specifications covering water works materials and their installation. These protective measures have been in use for the past five years and we have no doubt concerning their ef fectiveness, although a constant study and effort is being made to improve upon existing procedures and ma terials. In general terms the following measures, which have been approved by Management, are in current use: 1. Cast-iron pipe is given preference in San Antonio and is normally used in sizes ranging from 6-inch to 16inch, inclusive, in all cases where the resistivity is 700 ohm-cm or greater. Six-inch cast-iron pipe is also used in the corrosive areas, in which case it is embedded or encased in silica sand. Specifications for this sand require, among other things, that it contain not less than 98.5% silica, not more than 0.3% organic material and not more than 0.3% alumina (clay). Its electri cal resistivity as specified shall not be less than 25,000 ohm-cm at a moisture content of 5% by volume. 2. Asbestos-cement pipe for sizes 8-inch through 26-inch is normally used in the corrosive or "hot" areals! not normally made. Parallel local dis tribution mains of either cast-iron or asbestos-cement are installed in con junction with the installation of the trunk or feeder main. 4. The use of steel pipe for under ground installation is normally limited to the large diameter pipe required in the construction of primary pump sta tions. All such pipe, ranging in sizes from 24-inch to 54-inch, which has been installed during the past five years in connection with the construc tion of four new primary stations has been coated and wrapped with insulat ing material and cathodically protected. 5. Steel and cast-iron pipe used for installation in bored holes made for the purposes of crossing railways, high ways and other obstructions is normal ly primed and then wrapped with a 10-mil polyvinyl tape. Additional measures are required to prevent dam age to the protective wrap during in stallation. With an idea as to what the eco nomic limits were and a basis on which to proceed established, as far as the distribution system mains were concerned, other components of the system were looked into and corrective measures taken as required. These in cluded steel elevated and ground stor age tanks, meters and service lines, in cluding the customer yard piping. Class 200 pipe is specified for the 8- Summary inch asbestos pipe in order to obtain the added beam strength. In larger sizes, Class 150 is considered accept able. The fact that this pipe is a non San Antonio has a 'serious under ground corrosion problem which has resulted in excessively high mainte nance costs. conductor of electricity and relatively inert has resulted in greatly increased usage during the past five years. A test section involving some 5000 feet This problem has been given much study and a plan of attack has been developed and put into effect which includes the use of various protective of 6-inch Class 200 asbestos pipe was installed in a highly corrosive area this year. The performance of this pipe will be closely watched as to whether or not beam failure will occur due to measures for new construction, as well as a long-range replacement program. It is natural that any long-range project or plan of operation, however well executed, will have faults and de soil induced stresses upon the pipe. In all cases where asbestos cement pipe is used, the fittings, hydrants and valves are encased in silica sand. ficiencies here and there. These, however, will be discovered empirically as work progresses and re quire only a minimum of surveillance 3. Pre-stressed concrete-lined steel and care. The plan as a whole appears cylinder pipe is specified for 24-inch sound and it is believed that future through 36-inch sizes. This type pipe installation will eventually reach a high is gaining popularity due both to degree of perfection.