3
Corrosion Performance of Ductile Iron Pipe: Case Histories and Data

The corrosion performance of ductile iron pipe (DIP) in aggressive soils has been the subject of considerable debate, resulting in disagreement on the interpretation of publicly available data on DIP performance. The three underlying causes of the disagreement are these:

  1. Material failures in a system, in general, are due to a combination of factors, including variations in the material, installation, environment, and use, and thus often represent the tails of the distribution of behavior, not the averages, when compared with the behavior of most components in a field system or in a more controlled study of system components.

  2. Limited field corrosion performance and field failure data are available for water pipelines overall, not just for DIP.

  3. Data are reported in the form of averages, but water authorities need to make risk management decisions based on the data in the tail of the distribution.

The Bureau of Reclamation’s pipelines are long, cross many different types of terrain and environments, and are required to provide uninterrupted, reliable service. Reclamation’s corrosion control decisions have been made to limit the risk of premature corrosion failures while accommodating different types of pipe and environmental conditions. The report of the National Research Council’s Transportation Research Board entitled Transmission Pipelines and Land Use: A Risk-Informed Approach describes Muhlbauer’s analysis of the need for risk assess-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 43
3 Corrosion Performance of Ductile Iron Pipe: Case Histories and Data The corrosion performance of ductile iron pipe (DIP) in aggressive soils has been the subject of considerable debate, resulting in disagreement on the interpre- tation of publicly available data on DIP performance. The three underlying causes of the disagreement are these: 1. Material failures in a system, in general, are due to a combination of fac- tors, including variations in the material, installation, environment, and use, and thus often represent the tails of the distribution of behavior, not the averages, when compared with the behavior of most components in a field system or in a more controlled study of system components. 2. Limited field corrosion performance and field failure data are available for water pipelines overall, not just for DIP. 3. Data are reported in the form of averages, but water authorities need to make risk management decisions based on the data in the tail of the distribution. The Bureau of Reclamation’s pipelines are long, cross many different types of terrain and environments, and are required to provide uninterrupted, reliable service. Reclamation’s corrosion control decisions have been made to limit the risk of premature corrosion failures while accommodating different types of pipe and environmental conditions. The report of the National Research Council’s Transportation Research Board entitled Transmission Pipelines and Land Use: A Risk-Informed Approach describes Muhlbauer’s analysis of the need for risk assess- 

OCR for page 43
corrosion Prevention standards ductile iron PiPe  for ment for pipelines in the presence of “unmeasureable and unknowable” factors to “systematically and objectively capture everything that is known and use the infor- mation to make better decisions.”1 It is with this approach that the data described below have been collected and used as the basis for this report’s responses to the questions asked by Reclamation in the charge to the committee (see Chapter 1, the section entitled “Reclamation’s Request”). The committee reviewed the available data on the performance of DIP with polyethylene encasement (PE) and cathodic protection (CP) to assess the observed range of material and pipeline system behavior in aggressive soils. As a comparison, the committee also reviewed the available information on the performance of DIP without PE or CP, DIP with PE and without CP, and DIP with CP and without PE, as these data might represent local behavior if local failure of PE or CP in a field system was the underlying cause of observed corrosion and/or pipeline failure. Information was also reviewed for comparison with bonded dielectric coatings on both steel and ductile iron water and sewer pipelines. An additional factor that may come into play in a failure of operating pipelines is the design and installation of other components into the pipeline system. The committee was asked to use the failure data provided by the U.S. Depart- ment of Transportation’s (DOT’s) Office of Pipeline Safety (OPS) for operating steel pipelines with bonded coatings and CP as the benchmark standard for com- parison with data on DIP with PE and CP. Therefore, problems with design and installation failures that lead to external corrosion must be considered as contribut- ing factors to the performance of each system and cannot be discounted. The following sections present the findings of the data review based on presen- tations made to the committee; on publicly available documents; and on specific information provided by the Bureau of Reclamation and utilities and on additional data provided by other sources at the request of the committee. The obtained data on pipeline materials and operating pipelines fall into the following categories of data types. The data type classifications (1 through 5) are listed from the most quantitative to the most qualitative: 1. Data Type —Documented failures of operating pipelines due to external cor- rosion for pipelines of a gien pipeline type, including pipe thickness, pipeline age, and soil condition. Such information can be converted into a linearized maximum pitting or corrosion rate for that specific failure (assuming that pitting begins upon pipe installation) and into a failure rate per mile per year if the total number of miles of pipeline of that particular age and soil condition are known. 1 Transportation Research Board, Transmission Pipelines and Land Use: A Risk-Informed Approach (Washington, D.C.: The National Academies Press, 2004).

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories 5 of 2. Data Type —Measured maximum corrosion pit depths2 in a gien pipeline segment with no documented failure as a function of pipeline type, including pipe thickness, pipeline age, and soil conditions. Such information can be converted into linearized maximum pitting rates and, for measurements from a set of pipe segments, into a distribution of corrosion rates and a distribution of estimated failure times for a pipeline of a given thickness. 3. Data Type —Calculated means and standard deiations from collections of data of maximum pitting or corrosion rates of different pipe segments if and only if the input data fit the statistical model on which the means and standard deiations are based. If the data fit the model, then the means and standard deviations can be used to estimate the average corrosion rates in the tail of the distribution. For example, from the definition of a standard normal distribution,3 2.3 percent of samples are expected to have property values greater than two standard deviations above the mean. For a distribution of maximum observed corrosion rates that fits some distribution, therefore, 2.3 percent of pipeline segments in a system characterized by that distribu- tion would be expected to have maximum observed corrosion rates greater than two standard deviations above the mean. Similar types of statistical statements can be made for other types of distributions, for example, log- normal distributions. 4. Data Type —Calculated aerages of aerages or calculated mean alues assuming data distribution models, if the data do not fit the assumed statisti- cal models or if it is not known whether the data fit the statistical model on which the calculations are based. The averages are of limited use because the behavior at the tail of the distribution is what is needed for risk assessment. Means and standard deviations are useful only if the data on which the calculations are made follow the model. If the data do not fit such a model, then the calculated mean and standard deviation have limited physical sig- nificance. Data Type 4 may provide qualitative support for other sources of Data Types 1 and 2. If the data do not fit such a model, then the calculated mean and standard deviation have little physical meaning. If no informa- tion was provided about the validity of the model, the committee assumed that the data did not fit the model, based on the committee’s examination of similar data and their fit. 5. Data Type 5—Qualitatie data, including direct physical eidence of disrup- 2 Corrosion pit depth measurements are normally taken by examining the pipe surface in question to determine the deepest pit(s), which then are measured with a pit depth gauge. The measured pit depths are then divided by the burial time in years to arrive at a mil per year (mpy) corrosion rate. 3 The committee recognizes that there are other distributions against which these data can be measured. While the means would likely remain unchanged, these distributions (e.g., lognormal) may impact the tails.

OCR for page 43
corrosion Prevention standards ductile iron PiPe  for tion to external corrosion protection systems, such as damage to coatings or PE or failure of CP systems; indirect eidence of disruption to corrosion reduction systems by measured pitting rates or widespread external corrosion deiat- ing from that typically obsered; semi-quantitatie information on eidence of external corrosion and the frequency of disruption to corrosion reduction systems; and general descriptions of pipeline operator experience with specific pipeline systems where no quantitatie information is proided. The committee used Data Types 1 and 2 to calculate Data Type 3 as the primary basis for comparison with the OPS benchmark data using a frequentist defini- tion of probability in Chapter 4, which presents the failure criteria derived by the committee. OVERVIEW OF PRESENTATIONS TO THE COMMITTEE As noted in Chapter 1, at its first meeting the committee received several pre- sentations that provided it with a better understanding of the various perspectives and opinions regarding the committee’s statement of task. These presentations ranged from summaries of other research to informed opinion and are described in greater detail below. As summarized in Chapter 1 of this report, the first presentation to the com- mittee was made by representatives of the Bureau of Reclamation on its history, its concerns with the use of DIP with PE and CP, and the development of its Technical Memorandum 8140-CC-2004-1.4 A presentation was then made by L. Gregg Horn and his colleagues from the Ductile Iron Pipe Research Association (DIPRA)—an organization that represents the North American DIP manufacturers.5 Horn reviewed DIPRA’s history and described the 85 corrosion-related projects at 27 test sites involving more than 5,200 specimens that DIPRA has carried out since 1928. At present, DIPRA has 33 active corrosion studies under way at 16 test sites involving 2,000 specimens, with ongoing research projects with Corrpro, Schiff & Associates,6 and the Missouri University of Science and Technology (formerly, the University of Missouri-Rolla). Horn and colleagues stated that DIPRA’s empirical data prove that DIP with PE “works,” based on more than 150 field investigations of iron pipe with PE, either 4 Bureau of Reclamation Technical Service Center Staff, U.S. Department of the Interior, “Corro- sion Considerations for Buried Ductile Iron Pipe,” presentation to the committee, Washington, D.C., July 28, 2008. 5 L. Gregg Horn, DIPRA, PowerPoint presentation to the committee, Washington, D.C., July 28, 2008. 6 Corrpro and Schiff & Associates are corrosion engineering and consulting services headquartered in Medina, Ohio, and Claremont, California, respectively.

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories  of with or without CP, and that hundreds of millions of feet of iron pipe with PE are in service in the water and wastewater industry throughout North America. Horn and colleagues pointed out that DIPRA had conducted 11 field investigations on DIP with PE and CP on seven different pipeline systems and found no major cor- rosion. Richard Bonds, a statistician consulting for DIPRA, provided information and summary tables from DIPRA’s test data set on corrosion rates.7 This material included summary information from the test sites, field investigations, member company reports, and Corrpro reports for 1,379 individual data points. A more detailed discussion of DIPRA’s data set and the analyses thereof are provided in the section of this chapter entitled “Field Cases.” Horn then provided DIPRA’s perspective in answer to the two primary ques- tions comprising the committee’s charge. Regarding question 1—whether PE with CP works on DIP installed in highly corrosive soils—Horn stated that based on DIPRA’s research and extensive data sets, “polyethylene encasement works in highly corrosive soils without cathodic protection.” In response to question 2—whether PE and CP reliably provide a minimum service life of 50 years—DIPRA’s answer was “yes.” Its consultant for statistical analysis presented information and graphs of corrosion rates to support these positions.8 Horn and colleagues stated that DIPRA has always recommended against bonded coatings for DIP and that they are not aware of any North American DIP company which is a member of DIPRA that would provide DIP with bonded dielectric coatings. A presentation was given by Graham E.C. Bell of Schiff & Associates on the use of resistance probes and the success of the three cathodically protected pipelines (totaling 258 miles of DIP with PE and 68 miles of coated steel pipe) in North Dakota and South Dakota:9 (1) the Southwest Pipeline Project in North Dakota and (2) the WEB (Walworth, Edmunds, and Brown) Development Project and (3) the Mid-Dakota Rural Water Service Project, both in South Dakota. The presentation indicated that DIP wall thicknesses (t) will range from 250 to 480 mils depending on pipe diameter and pressure class. Bell pointed out that if the maximum general corrosion limit allowed is 50 percent of the pipe wall, then to reach a 50-year design life the maximum allowable corrosion rate would be 2.5 to 8.7 mils per year (mpy) depending on pipe wall thickness.10 Bell went on to state that for the pitting cor- rosion limit (through wall), the maximum allowable pitting rate would be 5.0 to 16.0 mpy. He concluded that CP must reduce or mitigate corrosion to these rates 7 Richard Bonds, L.M. Barnard, A.M. Horton, and G.L. Oliver, “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research,” AWWA Journal 97(6):88-98 (2005). 8 Charles Cowan, Analytical Focus, “Measurements and Standards,” presentation to the committee, Washington, D.C., July 28, 2008. 9 Graham E.C. Bell, Schiff & Associates, “Measurements of Performance of Corrosion Control Mechanisms on DIP,” presentation to the committee, Washington, D.C., July 29, 2008. 10 It should be noted that pressure in water pipes is lower than that in gas pipes.

OCR for page 43
corrosion Prevention standards ductile iron PiPe  for to “work.” He provided information on the historical use of electrical resistance probes for the Southwest Pipeline Project, where 20 steel probes had been installed at eight locations on the 76-mile-long DIP with PE and CP section of this 118- mile-long pipeline. He made the case that since these probes under the PE with CP indicated that the corrosion rate was less than 0.1 mpy, this confirms that CP “works” with PE. He stated that the Southwest Pipeline had only two problems with its corrosion protection systems. The first was due to high groundwater and the use of gravel bedding, which had damaged the PE in those areas and increased the CP current requirements so that a larger rectifier had to be installed. Excava- tions in one of these areas, even with potentials below protected criteria, indicated no corrosion on the pipeline. The second problem related to a leak on the 16-inch pipeline around Station 283+00 in the Southwest Pipeline Project. This leak is described in more detail in the section below entitled “Summary of Known Cath- odically Protected Polyethylene-Encased Pipelines.” Bell also discussed some of the results and problems that he and other consul- tants had observed with the use of certain steel probes on the Mid-Dakota Water Service Project. Bell and his colleagues found that some of the probes had failed because of manufacturing errors (inadequate seal of the back of the probe), which allowed water to enter the probe body; they concluded that these data could not be used because of faulty probe configurations. Bell also discussed development of a new ductile iron probe that theoretically should more closely represent corrosion rates on DIP.11 Another presentation was provided by Mike Woodcock on his experience as a metallurgist with corrosion and corrosion control methods for iron pipe.12 He dis- cussed the DIP microstructure, showed scanning electron microscope micrographs, and presented his experience with respect to widely varying corrosion rates seen for different DIP over the years. He believes that the corrosion rates are influenced by the chemical makeup of the iron pipe and the amount of trace metals. He said: [T]he older ductile iron pipes were made with a higher pig iron content (“pureiron”). The recent iron is predominantly made from scrap iron and steel. The scrap steel can contain chromium, nickel, molybdenum, tungsten, vanadium, titanium, etc. Some DIP being made today is extremely sensitive to corrosion; some is almost corrosion resistant. The presence of the trace metals and their carbides are believed to be a possible source of this wide variation in corrosion sensitivity. DIP is no longer an “iron product” with just Fe, Mn, Si, S, and P. Woodcock further explained how corrosion activity can be initiated at the 11 Bell, “Measurementsof Performance of Corrosion Control Mechanisms on DIP.” 12 Mike Woodcock, “Review of the Bureau of Reclamation’s Technical Memorandum,” presentation to the committee, Washington, D.C., July 29, 2008.

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories  of pipe surface when the asphaltic shop coating is damaged. He also discussed differ- ent types of coatings applied to DIP, including side and circumferential extruded polyethylene, tape coatings, coal tar polyurethanes, epoxy paints and filled epox- ies, ceramic epoxies, fusion bonded epoxy coating, PE, and European standards including sprayed zinc metal coating and zinc coatings with coal tar finish coats and PE. He requested that the committee help in allowing end users the option of deciding for themselves what type of corrosion control methods they want. He rec- ommended that an independent national database for all pipeline materials be set up to better track data on each type of pipe. He also asked that intelligent systems (smart pigs) be used on DIP with PE to provide additional information on how PE works. He noted that the effectiveness of PE in test pits and at test locations cannot provide the type of performance information that the end user needs. A presentation was provided to the committee by Michael Szeliga, a corrosion consultant, who stated that most corrosion occurs on 10 percent of a pipeline, with the most significant corrosion occurring on 10 percent of that 10 percent, so the chance of discovering the worst corrosion with a random test pit is only 1 percent.13 He presented several papers that discussed corrosion issues involving DIP with PE. These papers included references to 10 papers, authored either by the presenter or by others, that list problems with DIP and PE. This presentation was similar to the information that Szeliga provided in an earlier paper and presentation at the ASCE [American Society of Civil Engineers] 2008 Pipelines Conference.14 Szeliga con- cluded that, based on his experience, PE is likely to be damaged during installation regardless of the quality of construction, and the depth of corrosion under intact PE is about the same as the depth of corrosion where the PE is damaged. He also stated that, based on his experience, CP does not prevent corrosion under intact PE, and using external bonded coatings together with CP substantially reduces the number of pipeline failures. DATA COMPARISONS OF INTEREST The primary pipeline type of interest is DIP with PE and CP in highly corrosive soils—defined as soils having a resistivity below 2,000 ohm-cm. The uncertainty about the use of DIP with PE and CP under these conditions arises from various concerns: that CP may fail to be effective due to CP malfunction or electrical shield- 13 Michael Szeliga, Russell Corrosion Consultants, Inc., “An Independent Evaluation of the Effective- ness of Polyethylene Encasement as a Corrosion Control Measure for Ductile Iron Pipe,” presentation to the committee, Washington, D.C., July 29, 2008. 14 Michael Szeliga, Russell Corrosion Consultants, Inc., “An Independent Evaluation of the Effec- tiveness of Polyethylene Encasement as a Corrosion Control Measure for Ductile Iron Pipe,” ASCE Pipelines 2008 Conference, Atlanta, Georgia.

OCR for page 43
corrosion Prevention standards ductile iron PiPe 50 for ing, that undamaged PE alone may be insufficient to prevent corrosion in highly corrosive soils, that the PE is easily damaged and leads to localized corrosion as a result of holes or tears, and that installation issues, such as damage to the PE or corrosive soil adhering to the pipe remaining between the pipe and the PE during installation, change the local environment surrounding the pipe. The committee therefore performed additional assessments as estimates for the case in which one or both of the corrosion mitigation strategies fail—that is, assessments were made for DIP without PE or CP, for DIP with CP but without PE, and for DIP with PE but without CP. It is also important to note that all corrosion rates are reported and/or calcu- lated as linear corrosion rates, that is, the corrosion rate is constant over time. For Data Type 1, that is, for pipes that failed, a linearized maximum corrosion rate is calculated from the wall thickness of the pipe divided by the time to failure since installation. For Data Type 2, the measured maximum pit depths are converted into linearized maximum pitting rates by dividing the measured maximum pit depth by the burial time (years) since installation of the corroded pipe. Therefore, all pitting rates noted in this report are based on the assumption that the pit depth grows linearly with time. Corrosion is known to follow kinetics that are nonlinear, in many cases with an incubation time before corrosion begins. A square root of time dependence is often used. However, with the limited data available, the fact that most pitting rates are reported as linear, and with the wide variation of burial conditions that can change over time, a linear model is the simplest and the one commonly used. A study commissioned by DIPRA also made this same simplifying assumption, stating the following: Ideally, corrosion rate curves would be generated from the data obtained in this study and mathematical functions developed to predict realistic decreasing corrosion pitting rates for extended times of exposure. However, these functions vary not only with soil type but also with moisture, oxygen content, and bacterial counts, all of which can fluctuate over time. Additionally, the pipes in this study’s database were subjected to numerous soils, and these would have their own unique corrosion function. For this reason as well as for simplicity and conservatism, it was decided to treat the corrosion rate as a linear straight- line function. . . .15 Therefore, all of the corrosion rates cited in this report were calculated assuming that corrosion pit depths increase linearly with time. In addition, the terminology for pitting corrosion rates has been standardized to “maximum observed pitting rate” for Data Types 1 and 2, “mean maximum pitting rate” and “standard devia- tion of the mean maximum pitting rate” for data fitting a normal distribution, or 15 Bonds et al., “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research.”

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories 5 of “average maximum corrosion rate” for an arithmetic average for Data Type 3, and “mean maximum pitting rate” for Data Type 4. Cast iron and ductile iron corrosion rates were considered to be similar for purposes of comparison in the committee’s evaluation. Research by Romanoff of the National Bureau of Standards (NBS, now the National Institute of Standards and Technology) reported in 1964 and 1967 that ductile iron, cast iron, and steel corrode at similar rates in low-resistivity soils.16 After further testing, NBS reported in 1976 that ductile iron and steel “buried in the same soils . . . corrode at nearly the same rates when encased in some soils. Different soils, however, alter the corrosion rates for both materials.” 17 Based on the DIPRA data, Bonds et al. also concluded, “Overall results indicated that the corrosion pitting rates of ductile versus gray- iron pipe were soil specific to an extent but were essentially the same statistically (t-tests, 95% confidence). For this reason, the ductile- and gray-iron pipe data were combined to obtain the benefits of an increased sample size in subsequent analysis.”18 DIPRA indicates that including the cast iron in the ductile iron data set allows information on corrosion rates for older pipelines to be evaluated, since ductile iron has only been commercially available since the mid-1960s. In the DIPRA report summary, the condition of the PE was divided into two categories, either “undamaged” or “damaged.” “Undamaged” refers to the PE being intact at the location being examined and is a term that has generally been used in the field by most corrosion consultants, owners, and by DIPRA in its field summa- ries. “Damaged” PE may be defined as any opening in the PE that allows a change in the environment at the metal surface that would influence the corrosion rate locally for some easily observable distance from the damaged PE location. “Undamaged” PE in this report refers to having no observed physical damage directly over the pipe location where corrosion was observed and does not refer to any damage of the PE farther away from the corrosion pit that may influence the corrosion rate or environment at the corrosion pit. The DIPRA data sets of interest to the committee were those for DIPRA’s two most corrosive soil conditions, which would generally be under the Reclamation’s criterion of less than 2,000 ohm-cm. The committee assumed that these two DIPRA corrosivity soil designations would include the soils identified by the American Water Works Association (AWWA) in Appendix A of ANSI/AWWA Standard C105, 16 Melvin Romanoff, “Exterior Corrosion of Cast-Iron Pipe,” AWWA Journal 60(12):1129-1143 (1964); Melvin Romanoff, “Performance of Ductile Iron Pipe in Soils: An 8-year Progress Report,” presented at the AWWA Conference, Atlantic City, N.J., 1967; Melvin Romanoff, “Results of National Bureau of Standards Corrosion Investigations in Disturbed and Undisturbed Soils, Technical Bulletin No. 86,” presented at the Twelfth Annual Appalachian Underground Corrosion Short Course, 1967. 17 W.F. Gerhold, “Corrosion Behavior of Ductile Cast-Iron Pipe in Soil Environments,” AWWA Journal 68(12):674-378 (1976). 18 Bonds et al., “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research.”

OCR for page 43
corrosion Prevention standards ductile iron PiPe 5 for “Polyethylene Encasement for Ductile-Iron Pipe Systems,” as case 2 conditions (equal to or above 10 points) and case 3 conditions (uniquely severe soils).19 ANSI/AWWA Standard C105, Appendix A, defines the 10-point evaluation procedures and uniquely severe soils. In the 10-point soil classification procedure, soils are tested for five parameters (resistivity, pH, oxidation-reduction potential, sulfides, and moisture); if the assigned point values according to the table in ANSI/ AWWA Standard C105, Appendix A, add up to 10 or more, the soil is considered to be corrosive to iron pipe.20 Uniquely seere soils are defined in ANSI/AWWA Standard C105, Appendix A, as soils having the following three characteristics: (1) soil resistivity equal to or below 500 ohm-cm; (2) anaerobic conditions in which sulfate-reducing bacteria thrive—that is, neutral pH 6.5 to 7.5; low or negative redox potential, negative to +100 millivolts (mV); and the presence of sulfides (positive or trace); and (3) the water table intermittently or continually above the invert of the pipe. In order to evaluate the effectiveness of the different corrosion control meth- ods, the committee sought field data and supportive information from a variety of sources and compared the input received to determine answers to the following questions: • What is the corrosion rate of bare ductile iron or cast iron compared to iron pipe with PE? • Has corrosion occurred and at what rate under intact PE on ductile iron or cast iron as a comparison for locations where electrical shielding may occur? • How does the rate of corrosion compare under intact PE and intentionally damaged PE? • Is electrical shielding a concern for polyethylene-encased and dielectrically bonded coated pipelines and to what degree? • Has corrosion occurred and at what rate on DIP with PE and CP? • How do bonded dielectric-coated steel pipelines compare to bonded dielec- tric-coated DIP and to DIP with PE? In terms of the committee’s statement of work, note that a corrosion rate of 5 mpy for DIP with a wall thickness of 0.25 inch (6 mm) is a critical value in the simplified linear corrosion rate assumption, as a corrosion rate greater than this could lead to failure in less than 50 years. The limited amount of data available for this study by the committee was 19 American Water Works Association, ANSI/AWWA Standard C105/A21.5-05 (2005): “Polyethylene Encasement for Ductile-Iron Pipe Systems,” Denver, Colo. 20 ANSI/AWWA C105/A21.99, “Polyethylene Encasement for Ductile Iron Pipe Systems.”

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories 5 of compiled from various sources, including literature search, committee members’ experience, presentations to the committee during its first meeting in July 2008, and discussion and correspondence with other sources (consultants, material manu- facturers, applicators, and owners). The accuracy of the data provided for this evaluation could not all be independently verified because most of the data were provided by others from various evaluations and testing programs conducted over a number of years. However, every attempt was made by the committee to verify the data and to accurately summarize and present the data provided. BARE AND AS-MANUFACTURED IRON PIPE WITHOUT CATHODIC PROTECTION For comparison purposes, corrosion rates for bare and as-manufactured (shop- applied asphaltic coating) DIP and cast iron pipe (CIP) were reviewed with cases for which the PE was known to be damaged, installed incorrectly, or intentionally damaged. The reason that this is done for comparison purposes is that it is assumed that a properly operating CP system for these types of pipes would protect the lim- ited areas at damaged PE locations and thus exhibit low maximum corrosion rates. If the CP system is not working properly, then the maximum observed pitting rate may be similar to that in as-manufactured DIP (standard asphaltic coating) without CP and in DIP with damaged or intentionally damaged PE without CP. Further, if the shop-applied asphaltic coating is damaged locally, the maximum observed pitting rate may be similar to or higher than that in bare DIP without CP. Mean Maximum Pitting Rate for Bare and As-Manufactured DIP DIPRA provided information to the committee in the form of letters, reports, and articles on its 75-year test study with summary tables for sandblasted, bare, as-manufactured (asphaltic shop-coated), polyethylene-encased (undamaged or intentionally damaged), and vinyl-encased iron pipelines under a range of condi- tions.21 Exposure or burial times for the DIPRA ductile iron specimens ranged from 1 to 35 years and for gray iron from 1 to 103 years. This DIPRA study information presented to the committee in terms of “mean maximum pitting rates” is shown in Table 3-1, Rows 1 through 6.22 The mean maximum pitting rates (referred to by DIPRA as “mean deepest pitting rates”) were reported to have been calculated as the arithmetic average of the maximum pitting rates observed for individual pipe 21 Bonds et al., “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research”; Richard Bonds et al., “Corrosion Control Statistical Analysis of Iron Pipe,” Materials Performance 44(1):30- 34 (2005). 22 Cowan, “Measurement and Standards.”

OCR for page 43
corrosion Prevention standards ductile iron PiPe  for Southwest Pipeline One project in North Dakota on a 76-mile DIP line with PE and CP reported a major leak in October 2004.74 The leak occurred on this 12- to 19-year-old Class 250 (300-mil) walled pipe in the bottom of a small creek drainage. The CP designer stated in his presentation to the committee in July 2008 that he and his colleagues believed that the joint may have been overextended (deflected) and that either the joint leaked or that groundwater flowing through the low creek area locally increased the oxygen level and CP current requirements in that area. In 2004, the manufacturer of the ruptured DIP conducted a forensic analysis to evaluate the metallurgical and mechanical properties of the 16-inch-diameter sec- tion of pipe that had been returned to it.75 The inspection found that although the joint was overdeflected (6.6 degrees compared to a 5.0-degree design maximum), based on the manufacturer’s examination of the gasket and gasket sealing surface on both the spigot and the bell gasket recess area, it had not leaked. The manufac- turer also examined the thermite weld connections for the joint bond wires and found that the cadweld connection and mastic coating were in good condition and showed no evidence of galvanic corrosion. The pipe thickness in an undamaged area was measured and found to be greater (318 mils) than the specified Class 250 pipe thickness of 300 mils. The testing indicated that the pipe met the mechanical and metallurgical property requirements of ANSI/AWWA Standard C151.76 The manufacturer concluded that “the pipe showed no evidence of any type of leakage until it ruptured.” The pipe manufacturer also conducted additional testing and a scanning elec- tron microscope analysis of the leak area,77 concluding that the nature of the pipe wall failure was indicative of a catastrophic-type event as opposed to a leak over a long period of time. The manufacturer based this conclusion on the observations that the pipe was sufficiently thin that it ruptured due to stress overload; the failure area was jagged, which is indicative of a relatively recent event; the edges could fit back together completely; the surface of the breach exhibited less corrosion than the pipe surface; and the increased metal thickness on each end of the failure was sufficient to stop the rupture and handle the hydrostatic pressure. The manufac- turer described the external corrosion damage as a large area where the pipe wall 74 Bell, presentation to the committee, July 29, 2008. 75 American Cast Iron Pipe Company, Inestigation of the Fracture of a ” DIP from SWPP Station +: Scanning Electron Microscope Ealuation of Fracture (Birmingham, Ala.: American Cast Iron Pipe Company-Technical Division, October 11, 2004). 76 American Water Works Association, ANSI/AWWA Standard C151/A21.51-96, “American National Standard for Ductile-Iron Pipe, Centrifugally Cast, for Water” (Denver, Colo., 1996). 77 American Cast Iron Pipe Company, Inestigation of the Fracture of a ” DIP from SWPP Station +.

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories  of FIGURE 3-7 Location of 2004 leak on 16-inch Class 250 (300 mil thick) ductile iron pipe with polyeth- ylene encasement and cathodic protection in North Dakota. SOURCE: Courtesy of Joe Bichler, Bartlett and West Engineers. had been thinned down and stated, “The thinnest area consisted of a 2-[inch] wide band covering 120 degrees of the circumference adjacent to the face of the bell and the other area was about 1 sq. ft. at the very bottom of the pipe where the [pipe wall failure] occurred.”78 It concluded that the area of thin metal was due to external corrosion and that the rupture probably occurred when a transient pressure event took place in the water pipe system. The photo provided in the presentation of the leak is shown in Figure 3-7. Akron, Ohio The committee received information about a corrosion leak that occurred on a 16-inch DIP with PE and CP in Ohio in a less-than 6-year burial time.79 Histori- cally, a 7-mile section of CIP had experienced corrosion breaks in the same area. After a third leak occurred in the same area on the cast iron pipeline, a short, 1,300-foot section was replaced with Class 56 (520 mil thick) DIP with PE and galvanic anodes in 2001. Early in 2008 (after approximately 6 years of burial), the pipeline experienced a leak at a joint where the single joint bond was loose. From the information received by the committee, it is not possible to determine both whether that section of the pipe was the side with the anode and whether the joint 78 American Cast Iron Pipe Company, Inestigation of the Fracture of a ” DIP from SWPP Station +. 79 Gregg Loesch, Akron Public Utilities Bureau, communication and correspondence with the com- mittee, September 2008.

OCR for page 43
corrosion Prevention standards ductile iron PiPe 0 for FIGURE 3-8 Location of leak on bottom of cathodically protected 16-inch ductile iron pipe with poly - ethylene encasement in Ohio. SOURCE: Courtesy of Gregg Loesch, Akron Public Utilities Bureau. bond had been broken before excavation, during the pipe excavation, or as a result of pipe movement due to the pipeline leak. It was also not possible to determine the level of CP provided to the pipeline. The owner did confirm that the leak was on the bottom of the pipe and that there were no interference sources (rectifiers) in that area. See Figure 3-8. Vernal, Utah Three different types of DIP lines with PE (fire water, raw water, and sewer lines) were installed at a power plant in Vernal, Utah, in 1984, and CP was installed and operating by 1985.80 The raw-water and fire-water pipelines were excavated 80 William Spickelmire, discussions and correspondence with Jeff Mattson, Corrosion Control Technologies, Sandy, Utah, 1997 through 2008.

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories  of to repair broken test leads in 1995. Surface corrosion was found at two locations under intact PE, one on the raw-water line and one on the fire-water lines. The corrosion engineer, who observed the external corrosion damage, stated that “it was not from polyethylene-encasement coating defects, soil under the polyethylene, or [CP] stray currents at a broken joint bond.”81 The soil resistivities at these locations were approximately 3,000 ohm-cm. The estimated pit depths were 30 to 60 mils at one location and 125 mils at the second location and occurred in less than 10 years of burial on the raw-water and fire-water pipelines.82 In 2006, a DIP sewer line at this power plant was replaced because of six leaks.83 The corrosion engineer stated that it was not possible to determine whether the damage was due to internal or external corrosion or to a combination. The pipe was replaced and discarded before additional forensic testing could be performed to determine the actual depth of external corrosion. The engineer also reported that the pipeline displayed a thick black layer and severe graphitic corrosion, even with a measured protected potential of −1.10 V at the ground surface. Some of the corrosion damage may have been due to MIC activity, which indicates that CP may not be able to provide protection in cases of MIC if the PE is electrically shielding the pipeline. The sewer line case is included here for completeness but is not included in the failure analysis to be presented in Chapter 4. However, the maximum observed pitting rate for this line was 11.3 mpy, which is very similar to the 12.5 mpy maximum observed pitting rates on one of the water lines; the latter is known to be entirely from external corrosion. California City, California Corrosion was observed at three locations in California in 1983 on a section of a 14-inch DIP with PE and CP that had been removed because of a large leak.84 The pipeline potential measurements were slightly below adequate protection levels (−0.760 V potential range at the pipeline leak repair location) because of a short at an insulator on the opposite end of the pipeline. The 14- and 16-inch pipeline had been installed in 1975, and CP was added in 1979. The soils were extremely aggressive tidal mudflats with high chlorides (38,600 ppm) and a soil resistivity of 100 ohm-cm. The 14-inch DIP wall was perforated in three locations at the bot- 81 Spickelmire, discussions and correspondence with Mattson. 82 S pickelmire, “Corrosion Control Considerations for Ductile Iron Pipe—A Consultant’s Perspective.” 83 Spickelmire, discussions and correspondence with Mattson. 84 CH2M HILL, California City Corrosion Control Study, March 1983, with field observations by George Richards and William Spickelmire, (CH2MHILL: Engelwood, Colo.); California City Engi- neering Department, communication with the committee, September 2008.

OCR for page 43
corrosion Prevention standards ductile iron PiPe  for tom of the pipe near the spigot end of the joint. The pit that caused the leak was approximately 2.75 inches (69 mm) wide and 5.75 inches (144 mm) long. The other two locations where the corrosion pits had completely perforated the pipe wall were both approximately 1.5 to 2.0 inches (37 to 50 mm) in diameter. Only the cement mortar lining at these other two penetrations appeared to be holding the water pressure. The bond wires were tested and appeared to be functioning correctly, and the DIP was electrically continuous.85 It was not possible to determine whether the damage to the DIP wall had occurred prior to or after installation of the CP system (see Figure 3-9). Two other locations on this same pipeline were also excavated and examined, but no corrosion pitting was observed. Follow-up communication with the city’s engineering department revealed that this pipeline was abandoned and replaced with a new, 14-inch lined concrete cylinder pipe in 1987. SUMMARY OF BARE, AS-MANUFACTURED, AND POLYETHYLENE- ENCASED DUCTILE IRON PIPE CORROSION RATES This section presents published corrosion rates for bare, as-manufactured DIP and DIP with PE and derived corrosion rates for Data Types 1, 2, 3, and 4 in cor- rosive soils. It should be noted that the data presented are for both “undamaged” PE and “damaged” PE. The term “undamaged” refers to conditions in which there is no observed physical damage to the PE directly over the location on the pipe if corrosion has occurred. As a result of the PE damage, water or aggressive ions may enter at the damage site and influence the corrosion rate at a point away from the damage location. This general definition has been used by DIPRA and others and is also used in this report. Since some suggest86 that PE alone is ineffective because it cannot be constructed without damage, the committee chose to evaluate corrosion for DIP with both “undamaged” and “damaged” PE. Of specific note is that, for a more realistic comparison, data on both DIP and CIP were included. Stroud from DIPRA stated, “The number of documented failures of polyethyl- ene encased pipelines—the vast majority of which are the result of improper instal- lation—is insignificant compared to the miles of Cast and Ductile Iron pipe that are afforded excellent protection with this method of corrosion prevention.”87 85 CH2M HILL, California City Corrosion Control Study; California City Engineering Department, communication with the committee, September 2008. 86 Szeliga, “An Independent Evaluation of the Effectiveness of Polyethylene Encasement as a Cor- rosion Control Measure for Ductile Iron Pipe”; Spickelmire, “Corrosion Control Considerations for Ductile Iron Pipe—A Consultant’s Perspective.” 87 Troy Stroud, “Polyethylene Encasement Versus Cathodic Protection: A View on Corrosion Protec- tion,” Ductile Iron Pipe News (Spring/Summer 1998).

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories  of FIGURE 3-9 In California City, California, major corrosion damage on ductile iron pipe with polyethyl - ene encasement and cathodic protection with joint bond on opposite side of pipe. SOURCE: Courtesy of William Spickelmire, RUSTNOT Corrosion Control Services, Inc. However, a concern of Reclamation88 is that if corrosion occurs under undam- aged PE and application of CP is deemed necessary, electrical shielding of the CP currents to those areas that are experiencing corrosion may reduce the CP current below what is required to mitigate the corrosion. It is also possible that, even in the absence of shielding, CP may malfunction. Corrosion can occur at some distance from a break or flaw in the PE if oxygen is being replenished at a sufficient rate to continue the corrosion activity. This is similar to what has been observed on disbonded dielectric coatings. However, under anaerobic conditions, in the presence of sulfate reducing bacteria, MIC can occur under undamaged PE even when the pipe is under cathodic protection. MIC under disbonded pipeline coatings is a serious problem in the oil and gas industry and is not fully understood.89 Pikas points out that while CP can be effective at 88 Bureau of Reclamation Technical Service Center Staff, U.S. Department of the Interior, “Cor- rosion Considerations for Buried Ductile Iron Pipe,” presentation to the committee, Washington, D.C., July 28, 2008. 89 Joseph Pikas, “Case Histories of External Microbiologically Influenced Corrosion Underneath Disbonded Coatings,” NACE International Corrosion 1996, Paper No. 198, Houston, Tex.

OCR for page 43
corrosion Prevention standards ductile iron PiPe  for damaged coating locations, it becomes more limited when MIC activity occurs farther away from the damaged area, primarily due to the reduction of the current densities to those areas to provide adequate levels of CP effectively. Electrical shielding, defined as any insulating barrier that will prevent or divert the CP current from the structure that it was intended to protect,90 occurs if the disbonded coating or PE has sufficient dielectric strength to form a barrier that electrically restricts or isolates the pipe surface from the CP current. Although the CP current does not reach the pipe surface under the disbonded coating or PE, corrosion can still occur even when CP levels measure adequate at the ground surface. The amount of current reaching the pipe surface from a coating defect under either a disbonded coating or loose bonded coating such as PE is a function of the distance from the defect and the longitudinal resistance of the layer of soil or water between the insulating barrier and the pipe through which the CP current must pass to the active corrosion location. The smaller this annular space (separation distance) between the pipe and the insulating barrier, the higher will be the longi- tudinal resistance by unit length of the electrolyte (soil or water). This is because the reduced cross-sectional area of the electrolyte that will carry the protective CP current is smaller, and therefore there is a higher resistance per unit length. The distance that the current is able to penetrate from a coating defect or damaged PE location is therefore a function of the current density, the electrolyte, and the longi- tudinal resistance. This means that the ability of the protective current to penetrate small, annular spaces is limited. From a practical standpoint, according to Peabody, the distance that one can project current into a small space is approximately about 3 to 10 times the thickness of the annular space between the insulating barrier and the pipeline surface.91 If the insulating barrier is not an effective electrical insulator (e.g., if water absorption allows it to become conductive and current to pass through it, or if there are a number of damaged locations), then enough CP current may flow to the corrosion location to provide partial or complete protection. This is the principle behind perforated rock shield or microperforated PE (described later). Pitting Rates When looking at the pitting rate, it is of particular relevance to the committee’s work that if the rate is above 5 mpy for a pipe with a 250-mil wall thickness, then 90 A.W. Peabody, Peabody’s Control of Pipeline Corrosion, Second Edition, Ronald Bianchetti, ed. (Houston, Tex.: NACE International, The Corrosion Society, 2001). 91 Peabody, Peabody’s Control of Pipeline Corrosion, Second Edition.

OCR for page 43
corrosion Performance ductile iron PiPe: case Histories 5 of the desired 50-year pipeline design life would not be met.92 As noted by Bell,93 if a general corrosion limit of 50 percent of the wall thickness is the maximum allowed, then depending on the DIP wall thickness (250 to 870 mils), the maximum general corrosion rates to meet a 50-year life would be 2.5 to 8.7 mpy. He also points out that if the pitting corrosion limit of 100 percent of the wall thickness is the maxi- mum allowed, then depending on the DIP wall thickness, the maximum pitting rate to meet the 50-year design life would be 5 to 16 mpy. Table 3-10 summarizes these data sets. For the testbed samples in the DIPRA study,94 mean maximum pitting rates for soils ≥10 points and uniquely severe con- ditions as defined by ANSI/AWWA Standard C105, Appendix A,95 were included. However, for the other data sources no soil designation was reported. The DIPRA data used in conjunction with the Woolley report96 allowed the distributions of maximum observed pitting rates to be determined for the five pipe conditions reported in Bonds et al.97 For the other mean maximum pitting rates reported by DIPRA, the data distributions were not available to the committee, so the maxi- mum observed pitting rates could not be determined. It should be noted, however, that the DIPRA study mean maximum pitting rate for the uniquely severe soils with undamaged PE was 6.8 mpy, which is 15 times the comparable mean maximum pitting rate of 0.453 mpy for soils ≥10 points as defined by ANSI/AWWA Standard C105, Appendix A.98 Therefore, the maximum observed pitting rate of DIP with damaged PE should be much higher than the 8 mpy maximum observed pitting rate for undamaged PE in ≥10 corrosive soils. It should also be noted that the cor- rosion rate data are available as various data types—mean or average maximum (Data Types 4), maximum observed (Data Types 1 and 2), and minimum and maximum ranges (Data Types 1 and 2)—depending on the data source. In most cases, there were insufficient data provided to the committee to convert all of the mean maximum pitting rate data in the DIPRA study to the maximum observed pitting rate that is needed to compare various corrosion protection methods and to predict pipeline life. So although the data are not of the same types, the mean maximum pitting rates and the maximum observed pitting rates, where known, are included in Table 3-10 for reference purposes. 92 Bureau of Reclamation Technical Service Center Staff, presentation to the committee, July 28, 2008. 93 Bell, “Measurements of Performance of Corrosion Control Mechanisms on DIP.” 94 Bonds et al., “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research.” 95 American Water Works Association, ANSI/AWWA Standard C105 (2005), “Polyethylene Encase- ment for Ductile-Iron Pipe Systems,” Denver, Colo. 96 Woolley, letter and attached “Corrosion Database Statistical Analysis.” 97 Bonds et al., “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research.” 98 Bonds et al., “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research.”

OCR for page 43
 TABLE 3-10 Corrosion Rate Summary DIPRA Study Data with Mean DIPRA Study and Woolley Case Studies (Partial List): Pipe Maximum Pitting Ratea Analysis Report to DIPRAb Szeliga Datac From Summary Tables Row Conditions (Data Type 4) (Data Types 2 and 4) (Data Types 1 and 2) (Data Types 1 and 2) No data; assumed all pipes No data sought 1a Bare Table 3-1, soils ≥10 points, Table 3-2 soils ≥10 points, were as-manufactured 22 samples with mean of 15.1 22 samples with estimated (asphaltic shop-coated) mpy maximum observed pitting rate of 26 mpy 1b Table 3-1, soils uniquely See Note 1 severe, 173 samples with mean of 44.2 mpy No data sought Tables 3-1 and 3-2, varied 2a As- Table 3-1, soils ≥10 points, Table 3-2, soils ≥10 points, soils, 45 pipe examples 103 samples with mean of 103 samples with estimated manufactured with maximum observed 10.5 mpy maximum observed pitting (asphaltic pitting rate of 22.5 mpy. rate of 34 mpy shop-coated) Maximum observed pitting 2b Table 3-1, soils uniquely See Note 1 rates ranged from 3.182 severe, 70 samples with mean mpy to 22.5 mpy for of 28.7 mpy 45 pipe examples with measured pitting rates See Note 1 No data sought 2c Table 3-1, varied soils, 89 including 17 penetrations samples in five testbed sites with “combined mean” of 24.7 mpy for all five, all sites and the means for the individual sites ranging from 0 to 32 mpy depending on individual site Table 3-4, varied soils, maximum See Note 1 Tables 3-3 and 3-4, varied 3 Damaged PE Table 3-3, varied soils, 69 observed pitting rate of 68 mpy soils, 11 examples with samples in five testbed sites, for Szeliga site and 50 mpy for maximum observed pitting with “combined mean” of 11.2 Cape May, N.J., location. Maximum rate of 68 mpy. Corrosion mpy for all five sites, with the observed pitting rates ranged from rates ranged from 3 to 68 means for the individual sites 13.6 to 68 mpy for four pipelines mpy for 11 pipe examples ranging from 0 to 20.6 mpy with measured pitting rates depending on individual site including four penetrations

OCR for page 43
4a Undamaged No data; assumed all pipe Table 3-7, varied soils, maximum Table 3-5, soils ≥10 points, Table 3-5, soils ≥10 points, for PE 151 samples with mean of the 14 samples that showed examples had damaged observed pitting rate of 52.5 mpy. 0.453 mpy corrosion. Mean maximum PE, as there was no Maximum observed pitting rates designation in original ranged from 4.6 to 52.5 mpy for 14 pitting rate of 4.9 mpyd per Szeliga article summary examples with measured rates Woolley analysis; Table 3-6, chart for undamaged PE soils ≥10 points, 151 samples with calculated maximum observed pitting rate of 8 mpy 4b Table 3-5, soils uniquely See Note 1 severe, 85 samples with mean of 6.8 mpy 5 PE with CP No data No data No data Table 3-9, varied soils, maximum observed pitting rate of 86.6 mpy. Maximum observed pitting rates ranged from 3 to 86.6 mpy for six different DIP with PE and CP pipelines with measurable corrosion; pitting based on known 369 miles with 353 in soils below 2,000 ohm-cm (Table 3-8) of DIP with PE and CP. Four of the pipelines experienced leaks (with one sewer pipeline in Utah having six penetrations). The one pipeline in California was abandoned and replaced in 1987 after only 12 years of service NOTE: Table numbers refer to tables in this chapter of the present report. DIPRA, Ductile Iron Pipe Research Association; PE, polyethylene encasement; CP cathodic protection; DIP, ductile iron pipe; mpy, mils per year. NOTE 1: Insufficient data provided to convert uniquely severe soil mean rates or individual test site mean rate data to maximum observed pitting rates. aRichard Bonds, L.M. Barnard, A.M. Horton, and G.L. Oliver, “Corrosion and Corrosion Control of Iron Pipe: 75 Years of Research,” AWWA Journal , 97(6):88-98 (2005). bThomas Woolley, letter and attached “Corrosion Database Statistical Analysis” data presentation, April 7, 2008. cMichael Szeliga, “Analysis of Ductile Iron Corrosion Data from Operating Mains and Its Significance,” ASCE Pipelines, Advances and Experiences with Trenchless Pipeline Projects Conference, Boston, Mass., 2007. dU.S. Bureau of Reclamation, Materials Engineering and Research Laboratory, Technical Memorandum No. MERL-08-15, “Fountain Valley Project 2007, Security Lateral 16-inch Ductile Iron Pipe Failures,” U.S. Department of the Interior, Denver, Colo. (2007). 

OCR for page 43
corrosion Prevention standards ductile iron PiPe  for Data Evaluation for Ductile Iron Pipe with Polyethylene Encasement and Cathodic Protection in Soil ≥10 Points The maximum observed pitting rates for bare, as-manufactured, and dam- aged DIP with PE in corrosive soils was observed to range from 13.6 to 68.0 mpy, depending on the study. The maximum observed pitting rates for undamaged PE ranged from 4.0 to 52.5 mpy, depending on the study. These results indicate that, for whatever underlying reason (e.g., PE damage, damage to the asphaltic coating, MIC, and so on), DIP with PE installed in corrosive soils can be found to exhibit the maximum pitting rates characteristic of bare and as-manufactured DIP and DIP with damaged PE. The maximum observed pitting rates for DIP with PE and CP ranged from 3.0 to 86.6 mpy for the seven studies represented in Tables 3-1 through 3-10. These results indicate that, for whatever underlying reason (e.g., shielding, joint bonds, insufficient CP current, PE damage, damage to the asphal- tic coating, MIC, and so on), DIP with PE and CP in pipeline installations can be found to exhibit the maximum observed pitting rates characteristic of bare and as-manufactured DIP and DIP with damaged PE. These data indicate that DIP with PE and CP will not meet the target 50-year pipeline lifetime when installed in highly corrosive soils. For a nominal pipeline thickness of 250 mils, pipeline leaks can be expected in the worst case as soon as 3 years after installation. It should be remembered that lifetime is based on the behavior of the tail of the corrosion rate distribution, not on average values. There will be many pipe segments in a pipeline that will meet the target 50-year lifetime, but those in the tail of the distribution will not. The large number of high maximum observed pitting rates seen for this small number of total miles is indicative of the rates seen in the tail of the distribution and can be used to make risk management decisions.