Review of the Bureau of Reclamation's Corrosion Prevention Standards for Ductile Iron Pipe (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

6 Findings, Conclusions, and Recommendations The Committee on the Review of the Bureau of Reclamation’s Corrosion Prevention Standards for Ductile Iron Pipe reviewed and discussed the informa- tion and data presented to it and sought out by the committee in order to come to the findings, conclusions, and recommendations presented in this chapter. The information and data referred to above were classified as Data Types 1 through 5 in terms of their relevance to predictions of the first failures to occur in a pipeline system. Specifically, the first failures are expected to be represented by corrosion behavior at the tails of the distribution, where the corrosion is fastest, and not by average behavior. Data Types 1 through 3 as defined in Chapter 3 were therefore used as the primary basis for the findings, conclusions, and recommendations presented here. The charge of the committee included a responsibility to reply to two specific questions: • Question 1—Does polyethylene encasement with cathodic protection work on ductile iron pipe installed in highly corrosive soils? • Question 2—Will polyethylene encasement and cathodic protection reliably provide a minimum service life of 50 years? The committee charge went on to pose another alternative responsibility depending on the answers to questions 1 and 2: • Alternative Responsibility—If the committee answers either of the above 131 

132 Corrosion Prevention Standards for Ductile Iron Pipe questions in the negative, the committee will provide recommendations for alternative standards that would provide a service life of 50 years. With respect to question 1, after its review of the question, including oral and written clarifications from the Bureau of Reclamation, the committee finds that if manufactured and installed correctly, polyethylene encasement with cathodic protection provides a betterment to bare and as-manufactured ductile iron pipe without cathodic protection in highly corrosive soils. There is little controversy over the statement that polyethylene encasement (PE) alone will serve to provide an improved level of corrosion protection over bare or as-manufactured (asphaltic-coated) ductile iron pipe (DIP). Studies sponsored by the Ductile Iron Pipe Research Association (DIPRA) and cited in this report provide convincing evidence of this general result. These studies compared pitting rates in test sections of as-manufactured DIP buried in highly corrosive soils to rates for equivalent pipes with PE, also in highly corrosive soils. It should be noted, however, that the committee does not endorse some of the data evaluation methods used by DIPRA in these studies. That said, when comparing the reported mean maximum pitting rates of the 103 as-manufactured DIP samples to those of DIP with PE (Data Type 4), the DIPRA study results indicate that the mean maximum pitting rate of all 151 samples is decreased by a factor of about 23 (10.5 mils per year [mpy] compared to 0.5 mpy) compared to as-manufactured pipe without PE. According to histograms of this data set, the maximum observed pitting rates in the distributions dropped by a factor of approximately 4 (from 34 mpy to 8 mpy) by adding PE. The commit- tee found this to be convincing evidence that intact PE alone may offer significant corrosion protection to bare DIP. The committee believes that if cathodic protection (CP) is added to DIP with PE, evidence and experience from various sources lead to the conclusion that this may provide further improved levels of corrosion protection. While there is con- troversy as to whether CP can be relied on to protect DIP with PE at all distances from a defect or holiday in the PE, that does not detract from the finding that DIP with PE and CP may offer significant improvement in corrosion protection compared with bare and as-manufactured DIP (without PE or CP). In spite of the improvement obtained by the addition of PE and CP to DIP, the pipeline failures seen for DIP with PE and CP have occurred with maximum linearized pitting rates  Particularly troubling to the committee were the use of the weighted average or mean maximum pitting depths without maximum and minimum pitting depths, or the distribution of pitting depths reported, the combination of pitting data from various sites with very different corrosion behavior, the lack of reported time-dependent pitting depths (when such results are available), and unrealistically short burial times before excavation (as short as 1 year) combined with total study times as short as 3 years, for example, for the study of DIP with intentionally damaged PE. 

Findings, Conclusions, and R e c o m m e n dat i o n s 133 that are characteristic of bare or as-manufactured DIP in the same soils. From this information, the localized failures of the corrosion protection systems of PE and CP are inferred. However, for DIP with PE and CP, it is expected that the frequency of failure could be significantly lower than for bare and as-manufactured DIP without PE and for DIP with PE and no CP. With respect to question 2 in the charge to the committee, after its review of the question, including oral and written clarifications from Reclamation, the committee finds that the limited data available and the scientific understanding of corrosion mechanisms show that ductile iron pipe with polyethylene encase- ment and cathodic protection is not likely to provide a reliable 50-year service life in highly corrosive soils (<2,000 ohm-cm). In order to reach a finding with respect to question 2, the committee needed to know what threshold level of failures would be acceptable to Reclamation. In discussions with the committee, Reclamation expressed a desire to have no failures over the 50-year life of its systems because failures that interrupt service are very serious events. However, it is understood by Reclamation that it is never possible to engineer any system that will not have some probability of failure, and in answer to the committee’s asking Reclamation to define a level of failure risk that the bureau would like to attain for its pipeline systems, Reclamation responded by using data from the U.S. Department of Transportation’s (DOT’s) Office of Pipeline Safety (OPS) on welded steel gas pipelines with bonded dielectric coatings and CP to calculate a failure rate for that system of gas pipelines. Regarding such pipelines, Reclamation stated the following: “We have concluded that the level of performance provided by steel pipe, installed in severely corrosive soils with cathodic protection and bonded dielectric coating, is a reasonable benchmark (emphasis added) against which to measure the performance of ductile iron pipe installed in severely cor- rosive soils with polyethylene encasement and cathodic protection.” From the OPS data set, Reclamation calculated a failure rate of 0.000044 fail- ures per mile per year. The most important aspect of this failure rate is that despite many shortcomings considered and discussed earlier in this report and mentioned below, it did set a threshold level of risk tolerance for Reclamation. Perhaps one of the most obvious shortcomings of the calculated failure rate of 0.000044 failures per mile per year is that it does not consider the corrosivity of the soils. Reclamation noted this in its letter of August 21, 2008, with the statement: “However, the DOT database does not include information on the soil conditions in which the pipelines are installed, so we are unable to further screen the data to  Michael Gabaldon, Bureau of Reclamation, letter to Emily Ann Meyer, National Research Coun- cil, re: National Academies Review of the Bureau of Reclamation’s Corrosion Prevention Standards for Ductile Iron Pipe—A Response to the Committee’s Request for Clarification on Project Scope, August 21, 2008. 

134 Corrosion Prevention Standards for Ductile Iron Pipe include only pipe installed in severely corrosive soils. We are not able to quantify the impact this issue has on the calculated performance data noted above, but some adjustment to the computed failure rate may be warranted to compensate for this uncertainty in soil conditions across the data set.” The committee also noted early in its study that these data do not account for soil corrosivity. DIPRA has criticized the Reclamation benchmark on these grounds as well. The committee spent considerable effort in attempts to correct the calculated failure rates from the DOT OPS data for soil corrosivity. Searches for definitive data on the fraction of gas pipelines in highly corrosive soils (below 2,000 ohm-cm) were not successful. Attempts were made to estimate the fraction of soils that are highly corrosive and the fraction of failures that would have occurred in such highly corrosive soils where gas pipelines are located, but these attempts ended with the realization that these estimates were highly speculative and not well supported by any engineering data available to the committee. Therefore, the committee under- stands that the failure rate for pipelines buried in highly corrosive soils will likely be higher than the rate for the entire system, but it has no means of quantifying the degree of that likely increase. In further analysis of the rate of 0.000044 failures per mile per year calculated by Reclamation, the committee believed that the number of failures counted should be decreased since failures for pipe over 50 years of age should not be considered. By contrast, DIPRA criticized the number of failures as being too low because four failures in welded pipe seams due to external corrosion were not counted. The committee then counted those four failures and removed the failures in pipelines equal to or more than 50 years of age in a revised calculation. The committee also revised the total mileage of pipeline to recognize the total length of pipe, both failed pipe and pipe that did not fail. With these corrections, the committee calculated an additional threshold of about 0.000012 failures per mile per year. The pipeline data used for the Reclamation calculation included failures iden- tified between 2002 and 2008. DIPRA objected to that window as one in which stringent pipeline monitoring was in place for DOT pipelines, and it contended that if pipelines were monitored less stringently (as is generally the case for water pipelines), a higher failure rate would have been seen. The committee agreed with this contention and studied the gas pipeline data taken for pipelines before the current stringent monitoring guidelines were imposed. To establish an alternative threshold, a different calculation method was used, based on the failure data for external corrosion of gas pipelines before monitoring systems were imposed. The  Gabaldon, letter to Meyer, August 21, 2008.  Pipeline and Hazardous Materials Safety Administration, Code of Federal Regulations, Title 49, Part 192. 

Findings, Conclusions, and R e c o m m e n dat i o n s 135 committee calculated a threshold failure rate of 0.000041 failures per mile per year for gas pipelines 22.5 years old, the age characteristic of current pipelines of DIP with PE and CP in water systems. Both of the recalculated thresholds (0.000012 and 0.000041 failures per mile per year) are lower than Reclamation’s benchmark (and presumed risk-tolerance level) of 0.000044 failures per mile per year. The committee then sought to find failure data on DIP with PE and CP to compare to the benchmark. Only about 350 miles of such water pipeline that was buried in highly corrosive soils could be found. Much of these data are for Recla- mation projects, and within that set one failure was known. The committee also sought data from other pipeline users who were reported to use DIP with PE and CP. From these users, several other failures were reported. Some of these were for DIP with PE and CP used in a sewer application. The com- mittee recognized that, while these corrosion failures probably should not have occurred in a well-protected system, the application was sufficiently different from those used by Reclamation systems that this sewer pipe and its failures were not used in the calculations. Two other failures were reported to the committee in water-carrying DIP protected by PE and CP in highly corrosive soils. Both were on pipeline systems that were short compared to those of Reclamation but nevertheless were failures of water-carrying pipelines. Using the three failures referred to above on 353 miles of total pipeline, the committee calculated a failure rate for DIP with PE and CP. This was done two ways, both yielding about the same failure rate. In one method, a 23-year window of time was used (similar to the use of a time window for the OPS calculations), representing the reporting time for failures. This yielded a failure rate of 0.00037 failures per mile per year (see Eq. 4-4 in Chapter 4). The other method used a weighted average (weighted by the product of the time that the pipeline had been in service times the length of the pipeline) of the individual failures. This method is mathematically the same as considering the 353 miles of pipeline to have a weighted average age of 22.5 years. This method yielded a nearly identical failure rate of 0.00038 failures per mile per year (see Eq. 4-13). The committee found these failure rates, when compared to the Reclamation benchmark of 0.000044 failures per mile per year, to be sufficiently large (and even larger than the recalculated benchmarks from the committee) that it was concluded that DIP with PE and CP was not likely to meet the expectations of Reclamation for 50-year reliability or the risk-tolerance threshold set by Reclamation. This is particularly true considering that none of the pipelines studied were yet approach- ing the expected 50-year service life, and the number of failures is expected to rise further with age. Conducting two different sensitivity analyses based on three valid 

136 Corrosion Prevention Standards for Ductile Iron Pipe failures led to the same conclusion: DIP with PE and CP is not likely to meet the threshold set by Reclamation. The committee also reviewed pitting data on DIP with PE (without CP) in highly corrosive soils, as those data could be viewed as the conditions experienced by a pipe at a distance from a damaged area (holiday) in the PE where the pipe is shielded from the protective CP current. There was no consensus in the com- mittee on how to interpret such data, since the data are presented as linear pitting rates. Some committee members would argue that linearized pitting rates are a commonly used metric for anticipating failure, whereas others contend that since pitting is not expected to proceed at a linear rate with time, such analysis is not firmly rooted in good science. This disagreement might lead to the conclusion that only full data sets with the time dependence of pitting and corrosion up to failure may be used for analyses. For the purposes of this report, the observations of failures and the linearized rates that result are the only available information on DIP with PE and CP. The committee had to work with the data available. In doing so, it also recognizes that linear pitting rate data do not follow any accepted model of corrosion but nevertheless can give a conservative estimate of time to failure for comparison. This linear model is a common engineering practice and predicts a conservative estimate of time to failure, because actual corrosion that has an induction period and/or slows with depth of corrosion would result in actual lifetimes longer than those predicted by linear extrapolation. DIPRA has conducted burial and excavation studies of DIP in highly corrosive soils and shared limited aspects of its linearized data and its statistical analysis of the data with the committee. DIPRA studies involved burying 4- to 8-foot sections of bare DIP, as-manufactured DIP, DIP with intentionally damaged PE, and DIP with PE in corrosive soils in various locations and excavating sets of the pipes at regular intervals to evaluate maximum pit depth at each burial time. Mean maximum pitting rates were calculated on the basis of aggregate data for all burial times, and in many cases by aggregating information for multiple sites. The mean maximum pitting rate of DIP with PE are compared to pipes buried without the protection of PE, and the data as presented make a case for showing that mean maximum pitting rates are reduced by intact PE. The committee agrees with that qualitative conclusion. Of more interest to the committee were the actual data on observed pitting depths, but these data are considered proprietary to DIPRA and full details were not shared. However, DIPRA has reported (in Chapter 3 of the present report, see Table 3-5, Row 3) that of the 151 pipes (representing less than 1,000 feet of pipe), 14 exhibited measurable pitting corrosion and that the mean maximum pitting rate for these 14 samples is about 5 mpy. These data also suggest that the most extreme corrosion rate exceeded the mean of 5 mpy, and further analysis (see Table 3-6) showed maximum observed pitting rates that may be as much as 8 mpy. 

Findings, Conclusions, and R e c o m m e n dat i o n s 137 The committee believes that it is the extreme values (Data Types 1 and 2) that need to be considered rather than the reported mean maximum pitting rates (Data Type 4), because it will always be the point of most rapid corrosion on a length of pipe that will lead to the first failure. In this case, if there is a maximum observed pitting rate of at least 5 mpy, a linear extrapolation of that rate would suggest a corrosion depth of at least 250 mils in 50 years. Again, the committee does not accept that a linear maximum observed corrosion rate reflects scientific understanding of corrosion mechanisms, but such linear extrapolations are used in the absence of more comprehensive data. One DIPRA study reports a mean maximum pitting rate for this measurement on all 151 pipe sections as 0.45 mpy leading to a projected time to corrosion penetration of over 500 years. Using the mean maximum pitting rate (5 mpy) for the 14 corroding sample pipes, however, this projected time to failure would be only about 50 years. Anticipating that the maximum pitting rate for this distribution will be greater than the mean of 5 mpy and could be at least 8 mpy based on reported values, the projected time to first failure for this agglomeration of 151 short sections of pipe would be 32 years, well less than 50 years. Considering the DIPRA data referred to above on the pipe measuring between 600 and 1,200 feet (151 samples at 4- to 8-foot lengths) that is presumably installed with ideal care, the committee does not find that the studies of DIPRA confirm that DIP with PE can meet the expected reliability of over 50 years of service life. The committee recognizes that, based on an understanding of scientific principles, adding CP to DIP with PE will likely lead to an improved reliability over DIP with PE alone. However, the shielding effect of the PE at some distance from damage or a holiday suggests that in those areas the maximum observed pitting rates may not differ from DIP with PE only. Indeed, the oxygen depletion encountered in most of the pipe will protect it from much of the corrosion, but there remain questions as to whether this mechanism and the CP will protect the entire pipe. Therefore, the available limited data from the DIPRA studies do not lead the committee to the conclusion that DIP with PE and CP will necessarily have the required level of reliability over the 50-year service life. The committee also collected data on maximum observed pitting rates of DIP with PE from various other field studies, as summarized in Tables 3-7 and 3-9 in Chapter 3. The data in Table 3-7 for maximum observed pitting rates under PE frequently give linear pitting depth rates significantly greater than 5 mpy, indicat- ing that by conventional industry extrapolations, these pipes could fail in much less than 50 years and will not meet the Reclamation threshold in corrosive soils. Likewise field data, summarized in Table 3-9, for DIP with PE and CP provided data  Charles Cowen, Analytical Focus, “Measurements and Standards,” presentation to the committee, Washington, D.C., July 28, 2008. 

138 Corrosion Prevention Standards for Ductile Iron Pipe on actual wall penetrations and linear maximum observed pitting rates greater than 5 mpy. Therefore, these field data do not contradict the fundamental understanding of corrosion mechanisms which suggests that DIP with PE and CP will not offer the level of pipeline reliability desired by Reclamation. Finally, although this corrosion protection system should work in ideal condi- tions, fundamental scientific principles also indicate that there can be problems with the use of DIP with PE and CP. The shielding of the CP current at points away from damage or a holiday in the PE is a concern to many. Shielding prevents reliable calculation of anticipated corrosion rates and can make the normal monitoring systems used to show pipeline health less reliable. However, there continues to be disagreement in the pipeline community on the importance of current shielding. There are those who believe that shielding does make CP less effective and that corrosion mechanisms such as microbiologically influenced corrosion (MIC) are favored in the anaerobic environments when a food source is present underneath the PE. However, others who recognize these principles would argue that experience has shown them to be unimportant or not contributing to significant corrosion. With respect to the alternative responsibility in the committee’s charge, since the committee answered question 2 in the negative, it is asked by the Bureau of Reclamation to “provide recommendations for alternative standards that would provide a service life of 50 years.” After considerable study and deliberation, the committee finds that using the performance of bonded dielectric coatings on steel pipe with cathodic protection as a benchmark for reliability, and based on available information, it is unable to identify any corrosion control method for DIP that would provide reliable 50-year service in highly corrosive soils. This finding does not mean that any of the corrosion mitigation systems will not meet the required benchmark, but rather that the data are insufficient to draw a conclusion in either the affirmative or the negative. In arriving at this finding, the committee naturally looked at the corrosion con- trol method currently required in the Bureau of Reclamation’s Technical Memo- randum 8140-CC-2004-1 for highly corrosive soils—a bonded dielectric coating on DIP with CP. Indeed, there is evidence to support a conclusion that this method of protection is superior to that of DIP with PE and CP, but the data on each system are so insufficient that it is not possible to ensure that this system will provide cor- rosion protection for a service life of 50 years. This is particularly true in view of the fact that the benchmark of meeting the reliability of the gas pipelines is a high standard and one supported by years of experience with many miles of pipe. Also, the DOT system represents the state of the art in pipeline maintenance, including periodic studies of pipeline with technology such as intelligent in-line inspection (“smart pigging”), in which a remotely controlled measuring device is sent down the pipe to collect data such as pipeline wall thickness. The industry serving the water pipeline infrastructure seldom uses such advanced monitoring methods. 

Findings, Conclusions, and R e c o m m e n dat i o n s 139 The committee considered that a very simple argument might have led to a conclusion that DIP with bonded dielectric coatings could provide the desired level of reliability. That simple reasoning is that DIP, cast iron, and steel pipe are often thought to corrode at roughly the same rates. Therefore, if bonded dielectric coat- ings on steel pipe offer the desired level of reliability, bonded dielectric coatings on DIP would also provide the desired level of reliability. However, the committee could not rely on this reasoning because some of the committee members were of the opinion that coatings on DIP are bonded to a surface that is quite differ- ent from the surface of steel and that therefore there could be no assurance that the coatings would perform equivalently. Also, some members believed that the difference in the nature of the joints in steel and DIP might also affect the overall performance of the corrosion protection. Others thought that these coatings would provide performance similar to that of bonded dielectric coatings on steel pipe. The committee believed that adequate data were not available to fully endorse bonded coatings with CP for corrosion protection of DIP because of these uncertainties in the performance of bonded dielectric coatings on DIP. There is limited experience with bonded dielectric coatings for DIP in the United States, and some pipelines were installed as long as 33 years ago. To the committee’s knowledge, none of these pipelines has failed, but the length of pipe involved and the length of service for the pipeline prevent the committee from endorsing this as a method leading to the desired level of reliability. Beyond the lack of data for DIP with bonded dielectric coatings and CP, the principles used to understand the mechanisms that may prevent DIP with PE and CP from meeting the needed pipeline reliability criteria can in some cases be used to question the ability of DIP with bonded dielectric coatings and CP to meet the reliability benchmark. For instance, a criticism of PE has been that if the encase- ment is damaged, the CP will protect the pipe in the immediate vicinity of the damage but may not protect the pipe away from the damage due to shielding of the current. In principle, the same argument can apply to bonded dielectric coatings where, due to current shielding, CP may not protect a pipe from corrosion away from the damage if the coating is disbonded. Of course, not all damage to bonded dielectric coatings includes significant disbonding of the coating in the vicinity of the damage, so it is anticipated that DIP with bonded dielectric coatings and CP should perform better than PE with CP in these cases. This leads to a conclusion that DIP with bonded dielectric coatings and CP is likely to provide a higher level of protection to DIP compared to that provided by PE with CP, but this does not equate to a finding that DIP with bonded dielectric coatings and CP will meet Reclamation’s desired level of reliability. The committee considered other corrosion mitigation methods such as anti- MIC PE, microperforated PE, zinc coatings with epoxy or other types of top coats, controlled low strength material backfill, and building in a corrosion allowance. 

140 Corrosion Prevention Standards for Ductile Iron Pipe The committee finds that these other corrosion mitigation methods show prom- ise, but the evidence is too limited to make any recommendations for their use at this time. In summary, the committee finds that PE with CP can provide improved cor- rosion protection to DIP when compared to bare DIP or as-manufactured DIP in highly corrosive soils. However, the committee does not believe that DIP with PE and CP is assured to provide the level of reliability expected by Reclamation over the 50-year pipeline service life in highly corrosive soils. The committee con- sidered alternative corrosion control methods that would provide the desired level of reliability of DIP in highly corrosive soils. In view of the low level of experience on alternative systems, the committee cannot provide assurance that any corrosion control method for DIP will provide a reliable 50-year service in highly corrosive soils. Of the alternatives considered, the committee was of the opinion that because DIP systems with bonded dielectric coatings and CP that are in use are performing well, this corrosion control system for DIP was most favored. The data on the corrosion mitigation systems were limited for a variety of reasons, chief among them being limited sample size—that is, the total length of pipelines with DIP with PE and CP in corrosive soils is 353 miles, with a maximum age of 29 years. Unfortunately, significantly more data are not expected to be forth- coming other than some indicating additional experience on pipelines remaining in service. One promising way of protecting DIP in highly corrosive soils is to use bonded dielectric coatings with CP; it would be useful if more data on the efficacy of this protection method would also be forthcoming. Unfortunately, it is unlikely that significant amounts of DIP will be installed with bonded dielectric coatings and CP because the domestic DIP manufacturing industry will not supply DIP with bonded coatings and resists allowing customers to apply such coatings. Therefore, the prospect of amassing the data that would be needed to test the efficacy of bonded dielectric coatings with CP for DIP is limited (with the possible exception of DIP with bonded dielectric coatings manufactured outside the United States). Therefore, the committee concludes that making a more extensive analysis of the reliability of both DIP with PE and CP and DIP with bonded dielectric coatings and CP will not likely be possible in the foreseeable future. The committee recommends several areas in which additional studies could clarify its findings and provide data to improve the reliability of water-carrying pipe systems. • The committee recommends that to improve the reliability of existing pipelines, modern testing systems such as intelligent in-line inspection (“smart pigging”) should be studied and introduced in order to monitor more closely the corrosion of pipes and permit the repair of pipe systems prior to failure. 

Findings, Conclusions, and R e c o m m e n dat i o n s 141 The adoption of such pipeline monitoring systems is not common in the water industry, but if adopted, they could enhance the reliability of these pipelines, par- ticularly where failure would have large negative impacts. The method of random digging to diagnose the corrosion behavior of pipelines is inadequate to predict the state of corrosion for the total pipeline and thus will seldom identify the fastest corrosion rates at the tail of the distribution. Studies should be initiated to evaluate some of the promising corrosion control systems considered in this study for water pipelines in order to determine whether the adoption of any of these systems would permit the system to meet the desired level of pipeline reliability. • The committee recommends that data on pipeline reliability be assem- bled for all types of pipe specified by the Bureau of Reclamation in Table 2, entitled “Corrosion Protection Criteria and Minimum Requirements,” in Technical Memorandum 8140-CC-2004-1 along with the specified cor- rosion protection applied in the various soil types. While the challenge to Reclamation’s specifications concerned only DIP in highly corrosive soils, there are other categories of pipes (steel, pretensioned con- crete, and reinforced concrete) and corrosion mitigation methods (including mor- tar/coal tar epoxy, mortar, concrete, and concrete/coal tar/epoxy depending on the pipe material) that are not challenged, and implicit is the assumption that these pipe types and corrosion mitigation methods meet Reclamation’s expectations for reliability. With a quantitative analysis of the behavior of these pipeline types in the other two soil conditions, future consideration of benchmarks for needed reli- ability will be much clearer, and the appropriateness of the corrosion protection measures specified can be better understood. Also, the committee believes that Reclamation can enhance the reliability of existing systems by adopting some of the modern pipeline monitoring systems in use on natural gas pipelines. There is evidence that better monitoring has led to enhanced reliability on the gas pipeline system. It is expected that a similar asset management program for critical water pipelines would enhance the reliability of these systems as well.