development and testing of a linear polarization

Development and Testing of a Linear Polarization Resistance Corrosion Rate Probe for Ductile Iron Pipe

Web Report #4361

Subject Area: Infrastructure

Development and Testing of a Linear Polarization Resistance Corrosion Rate Probe for Ductile Iron Pipe

©2015 Water Research Foundation. ALL RIGHTS RESERVED.

About the Water Research Foundation

The Water Research Foundation (WRF) is a member-supported, international, 501(c)3 nonprofit organization that sponsors research that enables water utilities, public health agencies, and other professionals to provide safe and affordable drinking water to consumers. WRF’s mission is to advance the science of water to improve the quality of life. To achieve this mission, WRF sponsors studies on all aspects of drinking water, including resources, treatment, and distribution. Nearly 1,000 water utilities, consulting firms, and manufacturers in North America and abroad contribute subscription payments to support WRF’s work. Additional funding comes from collaborative partnerships with other national and international organizations and the U.S. federal government, allowing for resources to be leveraged, expertise to be shared, and broad-based knowledge to be developed and disseminated. From its headquarters in Denver, Colorado, WRF’s staff directs and supports the efforts of more than 800 volunteers who serve on the board of trustees and various committees. These volunteers represent many facets of the water industry, and contribute their expertise to select and monitor research studies that benefit the entire drinking water community. Research results are disseminated through a number of channels, including reports, the Website, Webcasts, workshops, and periodicals. WRF serves as a cooperative program providing subscribers the opportunity to pool their resources and build upon each other’s expertise. By applying WRF research findings, subscribers can save substantial costs and stay on the leading edge of drinking water science and technology. Since its inception, WRF has supplied the water community with more than $460 million in applied research value. More information about WRF and how to become a subscriber is available at www.WaterRF.org.


Development and Testing of a Linear Polarization Resistance Corrosion Rate Probe for Ductile Iron Pipe Prepared by: Emer C. Flounders, Jr. and Dale D. Lindemuth Corrpro, 7000 B Hollister Road, Houston, TX 77040 Jointly sponsored by: Water Research Foundation 6666 West Quincy Avenue, Denver, CO 80235 Water Environment Research Foundation 635 Slaters Lane, Suite G-110, Alexandria, VA 22314 and U.S. Environmental Protection Agency Washington, D.C. Published by:


DISCLAIMER

This study was funded by the U.S. Environmental Protection Agency (EPA), the Water Environment Research Foundation (WERF), and the Water Research Foundation (WRF) under Cooperative Agreement No. CR-83419201. EPA, WERF, and WRF assume no responsibility for

the content of the research study reported in this publication or for the opinions or statements of fact expressed in the report. The mention of trade names for commercial

products does not represent or imply the approval or endorsement of EPA, WERF, or WRF. This report is presented solely for informational purposes.

Copyright © 2015 by Water Research Foundation

ALL RIGHTS RESERVED. No part of this publication may be copied, reproduced

or otherwise utilized without permission.

Printed in the U.S.A.


v

CONTENTS

LIST OF FIGURES ...................................................................................................................... .vii

LIST OF GRAPHS ......................................................................................................................... ix

LIST OF TABLES .......................................................................................................................... xi

FOREWORD ............................................................................................................................... xiii

ACKNOWLEDGMENTS ............................................................................................................. xv

EXECUTIVE SUMMARY ........................................................................................................ xvii

CHAPTER 1: INTRODUCTION ....................................................................................................1 Project Approach .................................................................................................................2 Unfulfilled Project Objectives... ...........................................................................................2

CHAPTER 2: LPR AND PIPELINE CORROSION BACKGROUND..........................................5

CHAPTER 3: LABORATORY TESTING PLAN ..........................................................................9

CHAPTER 4: LABORATORY TESTING RESULTS AND DISCUSSION ...............................17

CHAPTER 5: CONCLUSIONS ................................................................................................... 37

CHAPTER 6: RECOMMENDATIONS....................................................................................... 41

REFERENCES ..............................................................................................................................43

ACRONYMS AND ABBREVIATIONS ......................................................................................45

APPENDIX A: Utility Questionnaire and Summary of Utility Questionnaire Responses ...........47

APPENDIX B: Aquamate™ Portable CORRATER® Instrument ................................................59

APPENDIX C: LPR Probe and Modifications ..............................................................................63

APPENDIX D: Laboratory Testing ...............................................................................................69

APPENDIX E: Field Data Collection Forms and LPR Probe Procedures ....................................77


vii

FIGURES

ES-1 Prototype field LPR probe with annealed ductile iron rings............................................ xviii

1 Participating firms across the country ....................................................................................4 2 Original LPR probe with ductile iron rings ...........................................................................9 3 Final LPR probe with DI rings.............................................................................................10 4 Side-by-side testing of LPR probes .....................................................................................12 5 DI thick plates ......................................................................................................................13 6 DI thick plates assembled in two-electrode soil box without soil .......................................13 7 Nilsson soil resistance meter connected to two-electrode soil box without soil ..................14 8 AquaMate meter connected to two-electrode soil box without soil ....................................15


ix

GRAPHS

1 Comparison of corrosion rate measurements probe 1 versus probe 2 ...................................20 2 Comparison of corrosion rate measurements average probe CR versus DI thick plates

CR ..........................................................................................................................21 3 LPR probes Corrosion Rate versus Imbalance ......................................................................23 4 LPR probe averages corrosion rate versus imbalance ...........................................................23 5 DI thick plates corrosion rate versus imbalance ....................................................................24 6 LPR probes corrosion rate versus calculated resistivity ........................................................26 7 DI thick plates corrosion rate versus resistivity .....................................................................26 8 LPR probe averages corrosion rate versus chloride content ..................................................30 9 DI thick plates corrosion rate versus chloride content ...........................................................30 10 DI thick plates saturated corrosion rate versus chloride content ...........................................31 11 LPR probe average corrosion rate versus pH ........................................................................31 12 DI thick plates corrosion rate versus pH ................................................................................32 13 Comparison of corrosion rate measurements bucket tests, lab tests and standard laboratory

tests ........................................................................................................................35 14 Comparison of resistivity measurements bucket tests, lab tests and standard laboratory

tests ........................................................................................................................35


xi

TABLES

1 Participating in-kind water utilities and trade associations ......................................................3 2 Revised soils testing plan .......................................................................................................11 3 Bucket testing corrosion rate data over time ..........................................................................18 4 Corrosion rate data (two-minute probe readings) ..................................................................19 5 Corrosion rate and imbalance data .........................................................................................22 6 Corrosion rate and resistivity data .........................................................................................25 7 Standard laboratory testing data .............................................................................................27 8 Effect of resistivity on corrosion ............................................................................................28 9 Corrosion rate, pH and chloride data .....................................................................................29 10 Comparison of corrosion rate data ........................................................................................33 11 Comparison of resistivity data ................................................................................................34


xiii

FOREWORD

The Water Research Foundation (WRF) is a nonprofit corporation dedicated to the development and implementation of scientifically sound research designed to help drinking water utilities respond to regulatory requirements and address high-priority concerns. WRF’s research agenda is developed through a process of consultation with WRF subscribers and other drinking water professionals. WRF’s Board of Trustees and other professional volunteers help prioritize and select research projects for funding based upon current and future industry needs, applicability, and past work. WRF sponsors research projects through the Focus Area, Emerging Opportunities, and Tailored Collaboration programs, as well as various joint research efforts with organizations such as the U.S. Environmental Protection Agency and the U.S. Bureau of Reclamation.

This publication is a result of a research project fully funded or funded in part by WRF subscribers. WRF’s subscription program provides a cost-effective and collaborative method for funding research in the public interest. The research investment that underpins this report will intrinsically increase in value as the findings are applied in communities throughout the world. WRF research projects are managed closely from their inception to the final report by the staff and a large cadre of volunteers who willingly contribute their time and expertise. WRF provides planning, management, and technical oversight and awards contracts to other institutions such as water utilities, universities, and engineering firms to conduct the research.

A broad spectrum of water supply issues is addressed by WRF's research agenda, including resources, treatment and operations, distribution and storage, water quality and analysis, toxicology, economics, and management. The ultimate purpose of the coordinated effort is to assist water suppliers to provide a reliable supply of safe and affordable drinking water to consumers. The true benefits of WRF’s research are realized when the results are implemented at the utility level. WRF's staff and Board of Trustees are pleased to offer this publication as a contribution toward that end.

Denise L. Kruger Robert C. Renner, P.E. Chair, Board of Trustees Executive Director Water Research Foundation Water Research Foundation


xv

ACKNOWLEDGMENTS

The authors wish to thank the participating utility partners, shown below, who assisted us during this program. Of special note is Tim Stevens of Monroe County Water Authority who provided soil samples and photos throughout the program. The authors would also like to thank the following individuals within Corrpro: Emer C. Flounders, PE, Co-Principal Investigator and author; Dan Crabtree, contributing author and field investigator; and Nancy Jacob, Manager of Corrpro’s Houston Soils Lab.

Organization City/State/Country City of Asheville Asheville, NC City & County of San Francisco San Francisco, CA Aqua Pennsylvania Bryn Mawr, PA Washington Suburban Sanitary Commission Laurel, MD City of St. Louis Water Division St. Louis, MO New Jersey American Water Voorhees, NJ Northern Kentucky Water District Erlanger, KY Monroe County Water Authority Rochester, NY Ductile Iron Pipe Research Association Birmingham, AL Louisville Water Company Louisville, KY Prince William County Service Authority Woodbridge, VA City of Sacramento Sacramento, CA Water District No. 1 of Johnson County Lenexa, KS Greater Cincinnati Water Works Cincinnati, OH City of Greensboro Greensboro, NC Metropolitan Utilities NE Omaha, NE

The authors would also like to thank the members of the Project Advisory Committee for their support and assistance:

Ryan Benner, Las Vegas Valley Water District Peter Gaewski, Retired, but formerly of Tata & Howard, Inc. Jeya Rajalingam, Sydney Water


xvii

EXECUTIVE SUMMARY

Linear Polarization Resistance (LPR) is an electrochemical method for measuring the corrosion rate of a material in a given environment. It is an indirect measurement of the corrosion rate of surrogate electrodes that are part of the testing apparatus. LPR has primarily been applied in aqueous environments, but it can also be used in soils. In soils, LPR measures the corrosivity of the soil to a specific metal. LPR has gained acceptance in Australia as a means of screening centrifugally-cast gray iron pipelines for areas of higher and lower potential corrosion. As practiced in Australia, soil samples are collected in the field from as near the pipeline as possible. The soils are then transferred to a laboratory where the LPR measurements are made in soil boxes under controlled moisture conditions.

An LPR data point provides a measurement of the corrosion rate at a certain time, for a specific soil/metal combination. The corrosion rate will vary over time at the same location, as field soil conditions change. Soil conditions often change in terms of moisture content, chemical content, and other physical and chemical variables. Corrosion products on the surface of the metal will also change over time. The corrosion rate will also vary both horizontally and vertically from any given point on a pipeline, as soil and surface conditions vary. Small chemical and physical changes in the environment can create a corrosion cell. These unpredictable changes in the natural environment make it desirable to collect more characterization data rather than less, at least until a generally accepted soil/pipeline corrosion model is developed that can accurately predict changes in corrosion rates.

This project focused on advancing and modifying the LPR technique for soils/pipelines by making it field-based, and eliminating or minimizing the need for laboratory analysis of the soils. These changes are expected to make the collection of useful data far easier and cheaper than the existing approach, thus significantly increasing the value of the LPR technique.

This project created a field LPR probe for the assessment of the corrosion rate of ductile iron pipe. No standard currently exists for LPR measurements in soil. A field probe needs to work in environmental conditions not controlled as they might be in a laboratory. The field probe could give LPR readings, but since this new probe was also applied in uncontrolled conditions and to a new metal (ductile iron), the data needed to be relatively consistent and repeatable, and verified in some way. The approach taken was to generate LPR data in association with other soil data in standard and non-standard conditions, allowing correlation of data from the field probe to standard soil analyses through multiple analyses. By demonstrating reproducibility of field data and correlating these data to soils data for which standards did exist, the relevance of field probe LPR data could be verified. Unfortunately, no actual field-testing of the field probes was conducted as originally envisioned, due to water utility coordination issues. Instead, all testing of the field probes was conducted in simulated field conditions in the laboratory with soil samples collected by the water utility partners from the field.

Primary accomplishments of this project were as follows:

1. The project team fabricated two identical prototypes of a field LPR probe for ductile iron pipe. The two field LPR probes included annealed ductile iron electrodes to customize these probes for the testing of external ductile iron pipe corrosion (Figure ES-1).


xviii

Figure ES-1 Prototype Field LPR Probe from this project with annealed ductile iron rings

2. Tested two identical, commercially available corrosion rate monitoring meters (AquaMate CORRATER meters) with the same probe to generate LPR data. These identical meters were found to generate comparable LPR data on field condition soils when used with the same probe. These results indicate data reproducibility regardless of which meter was used.

3. Tested two identical field LPR probes in simulated field conditions to generate LPR data on the same soil. These two probes were found to generate comparable LPR and conductivity data on field condition soils. These results indicate data reproducibility regardless of which of the two probes was used.

4. Determined that the field probe readings stabilized quickly once in contact with the test soils. Stable readings could be taken after two minutes of probe-soil contact time. Readings were taken over various timeframes, but shorter timeframes are desirable so that data generation can be maximized, and two minutes was found to be sufficient for stabilized readings under these test conditions.

5. Laboratory-tested the field LPR probes in a variety of soils that were collected from the field for the generation of LPR and other data, in particular resistivity/conductivity of the soils. The data generated by the field probes were correlated to other data through a series of tests in non-standard and standard conditions allowing comparison of the different types of data. Results of the testing include:

a. The field probe generated reproducible and consistent LPR measurements when tested on the same field collected soils.

b. The field probe-generated LPR data were compared to LPR data generated by the soil box method. Good correlation between the two sets of data was found, but the soil box LPR data (corrosion rate) were typically higher than that measured by the probe.


xix

c. The field probe-generated LPR data were compared to LPR data generated by the soil box method under typical test conditions of saturated soil. No standard exists for generation of soil LPR data, but saturated soil conditions are a standard test condition set forth for generation of soil resistivity data (ASTM 2005) and would theoretically seem to correlate with LPR measurement methods in aqueous solutions. Good correlation between the two sets of data was found, but the soil box LPR data (corrosion rate) in saturated soil conditions were almost always higher than that measured by the probe.

d. The field probes generated reproducible and consistent soil conductivity measurements on field condition soils. Soil resistivity is the reciprocal of soil conductivity, and thus can be calculated from soil conductivity.

e. The field probe-generated soil conductivity data were compared to soil conductivity data generated by the soil box method. Good correlation between the two sets of data was found.

f. The field probe-generated soil conductivity data were compared to soil resistivity data generated by the soil box method under standard saturated soil conditions using the Nilsson resistivity meter (ASTM 2005). Good correlation between the two sets of data was found both in terms of absolute measurements and in terms of data trends.

The soil samples collected and analyzed in this project were not correlated with detailed

ductile iron pipe corrosion data, so which set of LPR corrosion rate data are most correlated with ductile iron pipe corrosion in the field cannot be determined. However, based on the data generated in this project, it seems that adequate proof-of-concept testing has been completed to verify that a field-based LPR probe could be developed and its data correlated with corrosion losses from ductile iron pipe. A field-based LPR probe would allow generation of more LPR (corrosion) data more quickly, and these data would be helpful in screening existing ductile iron pipelines for areas of higher corrosion rate.

BACKGROUND

The North American water and wastewater community has hundreds of millions of feet of

ductile iron pipe in service. Only a portion of the inventory has any form of external corrosion control. Ductile iron pipe, in certain environments, is subject to external corrosion.

Considerable research has been done on soil characteristics likely to result in corrosion of metallic pipelines. However, despite much work, no single unifying model of soil corrosion has been developed that is generally accepted. What is generally acknowledged is that corrosion rates will vary over time and distance (Romanoff 1957, Ricker 2010, Cole and Marney 2011, Rajani et al. 2011). Clearly, a complex set of chemical and environmental variables impact pipeline corrosion.

LPR is an electrochemical method for measuring the corrosion rate of a material in a specific environment. A detailed description of LPR theory and application in soils can be found in a recent Department of Transportation study (Silverman 1996, Farrag 2010). LPR measures the electrochemical resistance of the surface in its environment. The lower the measured polarization resistance, the higher the general corrosion rate. In addition, the Current Imbalance between the


xx

electrodes can also be measured. When the Current Imbalance is higher, there is a greater tendency for localized, or pitting, corrosion (Farrag 2010).

As applied to pipelines, LPR provides an indirect method of assessing the instantaneous corrosion rate of a ferrous pipeline. The LPR technique does not require the pipe to be taken out of service. Nothing has to be inserted into the pipe, nor, theoretically, would the pipe need to be excavated for direct measurement of corrosion losses. The existing LPR technique, primarily used in Australia where it is available from a commercial vendor, involves the collection and removal of test soils to the laboratory, where the soils are carefully prepared and LPR measurements are made in a soil box. The resulting LPR data provides information on the corrosion rate of the pipelines, and these data are useful in prioritizing existing pipelines for further investigation. Some utilities, especially Sydney Water and Hunter Water in Australia, have found LPR to be a cost-effective screening method to assess their centrifugally-cast gray iron pipelines, and have used the technique year after year (Dafter 2014). This LPR technique, as applied in Australia, has not been developed for application to ductile iron pipe, nor has it been developed for field generation of LPR data. Development of a field-based LPR probe for ductile iron pipe may provide more data that could be useful in preliminary assessment of buried ductile iron pipelines.

APPROACH

A prototype LPR field probe had already been developed by the researchers in previous

work, but modifications were envisioned to improve the probe’s usefulness in the field. The probe also needed further adaptation for improved measurement of corrosion rates on ductile iron pipes. Thus, two identical prototype field LPR probes were created as part of this project. Improvements on the original prototype included a smaller diameter body, relocation of wiring, and use of heat-treated ductile iron rings in the probe with an annealing surface oxide, similar to that found on ductile iron pipe.

Utilities were asked to provide three soil samples from their service area to be characterized and tested with the new LPR probes in the laboratory. These soil samples were used to generate LPR data from the new probes, and the data were correlated to other soil characteristics associated with soil corrosivity. All work was done in the laboratory, with simulated LPR field data generated by testing soils in buckets, and correlating these data with soil box and other testing done on the same soils.

RESULTS/CONCLUSIONS

This project established confidence that measurements from the two identical prototype

LPR probes were similar in the same soil, were reproducible, and are indicative of soil corrosivity. In soil, the probes reached equilibrium and stabilized within two minutes. Side-by-side testing of the probes in the same soils obtained comparable corrosion rate values. The LPR data also trended well with the data obtained using an established two-electrode soil box corrosivity measurement. Also, solution conductivity readings taken by the LPR probes and converted to soil resistivity data were in close agreement with standard laboratory-generated soil resistivity data. Further studies will be required to establish a direct correlation between field-based LPR measurements (corrosion rate), actual corrosion loss, and possible pit corrosion of in-service ductile iron water mains.


xxi

APPLICATIONS/RECOMMENDATIONS

A fully developed field-based LPR corrosion rate probe for ductile iron pipe could be an important tool for the water industry. Such a tool would allow many measurements to be made along the alignment of an existing pipeline. The sturdy construction of the probe developed in this project allows it to be pushed into the soil to obtain soil LPR and related data. By providing instantaneous direct readings of the corrosion rate at a specific site, the timeframe for complete corrosion penetration of a section of ductile iron water main could be projected based on those readings. Depending on the nature of the data, a more detailed examination of a pipeline could be better prioritized, or subsequent follow-up readings could be scheduled. A large quantity of LPR data along a pipeline would allow more accurate predictions of corrosion rate and better predictions of corrosion penetration of the pipeline. While this project has demonstrated a viable field probe, more research will be required to determine key probe reliability issues and fully develop the empirical model to relate these LPR field measurements to corrosion losses and also possibly corrosion pit depth of ductile iron pipelines. Future studies to support development and eventual commercialization of this probe would need to include detailed soils testing near existing pipelines and excavation of pipeline sections to allow direct readings of pit depths and overall corrosion loss. These data could then be related to the LPR field measurements.

RESEARCH PARTNERS

U.S. Environmental Protection Agency (EPA) Water Environment Research Foundation (WERF)


1

CHAPTER 1 INTRODUCTION

Within the water and wastewater industry, ductile iron pipe is widely used both in service and for new installations. Since its introduction in the 1950’s, North American utilities have installed hundreds of millions of feet of ductile iron pipe with only a portion of that pipe having a bonded coating or polyethylene (PE) encasement for corrosion prevention (NAS 2009). The corrosion resistant characteristics of the annealing oxide and asphaltic coating included in the standard manufacture of ductile iron pipe minimize the external corrosion attack of this product in many soil environments. However, ductile iron pipe is not immune to external corrosive attack from the environment. Similar to other metallic pipe, ductile iron corrodes in a complex process from various causes. These include bimetallic corrosion cells, differential concentration corrosion cells, acidic soils, highly conductive soils, wet soils, soil contamination from road deicing salt, stray direct current interference from nearby cathodic protection systems and direct current powered rail transit systems, and microbiological activity, amongst others.

Ductile iron main failures caused by external corrosion are related to the corrosion rate, wall thickness, and time of exposure. Given the introduction of ductile iron pipe in the 1950’s, the consequences of failure are becoming ever more prevalent in older water and wastewater systems, and especially those that are understood to be exposed to high soil corrosivity. The AWWA report, “Buried No Longer,” states that, “More than a million miles of pipes are nearing the end of its useful life and approaching the age at which it needs to be replaced” (AWWA 2012). As defined by Marlow et al., “End of asset life is the time at which a significant (capital as opposed to operational) investment is made. Remaining life is the time left before a significant capital intervention is required” (Marlow et al. 2009). While considerable attention has been given to this matter over the last several years, there remains a strong need for utility purveyors to have a straightforward tool to cost effectively and reliably assess their existing ductile iron infrastructure and predict the remaining useful service life of these facilities. This very prevalent need sets the foundation for the subject research.

LPR is an indirect method of assessing the corrosion rate on a given ferrous pipeline as compared to excavation of the pipe and direct measurement of pit depth and corrosion loss. The LPR technique has been primarily used in Australia applied to centrifugally-cast gray iron pipelines by a commercial vendor. As the practice is understood, LPR data on corrosion rates are used as input to several proprietary Weibull-based algorithms that predict corrosion loss, pit depth distribution, and pit penetration of a pipeline. The existing LPR technique involves removal of test soils to the laboratory where the soils are carefully prepared and then LPR measurements made in a soil box. The collected soils are obtained from locations as close as possible to the existing pipelines, both in terms of horizontal and vertical proximity. While non-destructive, and applicable to pipelines still in service, the removal of the soils from the ground and transfer to the laboratory to generate LPR data causes considerable difficulty, cost, and delay in data generation. Currently there is no commercially available field tool that can cost effectively and reliably assess the corrosion rate of existing pipe. Development of a field-based LPR probe has promise to provide data in a fast and easy manner that could be useful to assessing buried pipelines.


2

PROJECT APPROACH

This project was focused on development and proof-of-concept testing of a field-based LPR probe. A field-based LPR probe had already been developed, but modifications and improvements to that original probe were anticipated to make it more useful in the field, and to make it more specifically applicable to ductile iron (DI) pipes. Thus, machining and fabrication of the new LPR probes for DI pipe was a critical part of this work. Once fabricated the new prototype LPR probes could be used to generate LPR data.

In order to support the testing of the new prototype probes, the research team requested soil samples from participating utilities to serve as a test medium. The soil samples were to be collected from known areas variable soil conditions ranging from non-corrosive to very corrosive relative to pipeline corrosion. The soil samples would be used to confirm reproducibility of data from different soils obtained by different LPR corrosion rate probes. The same soil samples would be characterized in the Houston Soils Laboratory. In addition, the project team would perform comparative laboratory LPR corrosion rate measurements using a “soil box” (also used for soil resistivity) with ductile iron electrodes and the same soil samples to verify that the new prototype LPR probes were accurate.

UNFULFILLED PROJECT OBJECTIVES

The original goal of the research was to present industry with a technically sound, well-proven and user-friendly methodology for non-destructively determining the current condition and predicting the longevity of unwrapped ductile iron pipe. The project team believed that the development of an empirical model relating soil characteristics to pipe corrosion attack and corrosion deterioration rate was feasible and attainable. With such an empirical model in hand, practitioners could cost effectively evaluate existing ductile iron pipe infrastructure and prioritize areas relative to capital and maintenance expenditures impacted by external corrosion degradation.

The original planned approach was to engage participating utilities to gather soil data from pipe excavation sites and measure key parameters on problematic DI pipe when exposed for repairs. Data to be generated were expected to include pit depth, pit volume, and pit density (distribution) as well as collection of soil samples from these same excavation sites. It was also expected that some of the participating utilities had already generated a great deal of these types of data that could also be collected and used in this project. These utility-provided data were important to provide a basis for developing models of corrosion loss and pit corrosion propagation. In addition, the research team already had developed a prototype field-based LPR probe in earlier work that would be used to measure LPR data from field excavations as well as soil samples in the lab. Correlating these LPR data with pipe age and pit measurements, a projection of remaining time to full pipe wall corrosion penetration could be made and correlated to LPR measurements.

In developing this project, the research team had received acknowledgements and commitments from fifteen water and wastewater utilities and one trade association to provide in-kind contributions and data to this project. These utilities are presented in Table 1 and their distribution across the country presented in Figure 1.


3

Table 1 Participating Utilities and Trade Associations

1. Aqua Pennsylvania - Bryn Mawr, PA 9. Louisville Water Company - Louisville, KY

2. City of Asheville - Asheville, NC 10. Metropolitan Utilities District - Omaha, NE

3. City of Greensboro - Greensboro, NC 11. Monroe County Water Authority - Rochester, NY

4. City of Sacramento, Department of Utilities - Sacramento, CA

12. New Jersey American Water - Voorhees, NJ

5. City of St. Louis Water Division - St. Louis, MO

13. Northern Kentucky Water District - Erlanger, KY

6. City & County of San Francisco - San Francisco, CA

14. Prince William County Service Authority - Woodbridge, VA

7. Ductile Iron Pipe Research Association - Birmingham, AL

15. Washington Suburban Sanitary Commission - Laurel, MD

8. Greater Cincinnati Water Works - Cincinnati, OH

16. Water District No. 1 of Johnson County - Lenexa, KS

Part of the scope of work was to gather and prepare a profile of the fifteen participating

water and wastewater utilities through a questionnaire. The responses helped the project team learn about each utility partner including: their total length of system pipe and what materials comprise it; their customer base; number of failures; availability of leak records or existence of a corrosion control program; and if they had the ability to collect soil samples and obtain specific pipe data at open trenches and exposed pipes. The survey form and the compiled responses are provided in Appendix A. The questionnaires revealed that the participating utility engineering and operations personnel did not have extensive corrosion and soils data from past failures, and would not be able to coordinate access to actual excavation sites nor gather any field data themselves. The budget was not adequate to provide the tools needed by the utilities to generate the needed data. Also, the timeframe of the project was not adequate to allow generation of a large quantity of data upon which to base the models of pit growth and propagation.

The expectation had been to use corrosion data of various sorts, especially corrosion pit data, supplied by utilities, as key criteria in the development of an empirical model for failure, correlating those data with field-generated LPR data. However, given the general lack of pit information, the project focused on further development and proof-of-concept testing of the prototype field-based LPR probe.


4

Figure 1 Participating utilities and trade organizations across the country


5

CHAPTER 2 LPR AND PIPELINE CORROSION BACKGROUND

This project developed an LPR probe for field use. The probe would be used along a

ductile iron pipeline to take measurements on the expected corrosion rate of a “bare” (no polyethylene encasement) ductile iron pipe. LPR is an electrochemical technique that measures the corrosion rate of a given material in a given environment. LPR is an indirect method of assessing instantaneous corrosion rates, and has primarily been used in aqueous environments to assess corrosion potential (Silverman 1996, Benjamin et al. 1996, Kirmeyer et al. 2002, ASTM 2014). LPR has also been applied in soils in a more limited manner for assessment of water pipelines, and in this situation, the soil moisture is providing the environment/electrolyte that allows for corrosion. When applied to soils and the assessment of pipelines, it has the advantages of the pipeline remaining in service while being assessed, and the pipeline does not need to be excavated. Since LPR gives an instantaneous corrosion rate, further interpretation of the data are needed to predict corrosion penetration of a pipe wall. The speed of generating LPR data is a key advantage of this technique in assessing corrosion when compared to coupon sampling /coupon corrosion analysis (Dafter 2014).

Electrochemical corrosion only requires an anode, a cathode, an electrolyte, and a circuit to connect the anode and cathode. At the anode, the metal corrodes putting metal ions into solution. This reaction releases electrons that pass through the metal of the pipeline to the cathode. At the cathode water and dissolved oxygen react to absorb the electrons and to produce hydroxyl ions in solution. The surface of the metal comprises a number of small anodes and cathodes, which form local corrosion cells. These local cells form from small differences in the metal or solution (Rohrback Cosasco Systems Inc. 2014). These same considerations apply to corrosion of a pipeline, making assessment of the corrosion rate on a pipeline extremely complex.

Considerable effort has gone into trying to understand the corrosion of buried pipelines and the corrosion characteristics of soil. Although much experience has been gained in regards to installation and failures of metallic pipelines, this experience has not lead to the development of a comprehensive and generally accepted model of pipeline corrosion in soils. Development of a model of corrosion in soils is the focus of ongoing research, including ongoing collaborative work by the Water Research Foundation. Corrosion in soils is a complex process influenced by soil and environmental factors including salinity and pH levels, moisture levels, degree of oxygenation, grain size distribution of the soils and related pore spacing, the wetting of the pipe surface by moisture in the soil, the formation of oxide or other layers on the metal surface, the presence of chlorides, the presence of sulfides, the presence of microbes that may facilitate corrosion, and potential stray currents (Cole and Marney 2012, AWWA 2010). The variables that influence corrosion in soils vary in both space and time.

Romanoff conducted considerable work on the condition of buried pipelines and relating these conditions to soil characteristics. He considered corrosion in soils from the basics of corrosion theory. For electrolytic effects, he highlighted pH, concentration of soluble salts and the occurrence of chemical reactions between the corrosion products and the electrolyte. Romanoff indicated that soil resistivity was a useful measure of the dissolved salts in the soil, and that this measurement appeared to correlate with corrosion in that soil (Romanoff 1957). However, while some studies have found correlations between some soil parameters and pipeline corrosion, other studies have found weak or no correlation (Cole and Marney 2012). For instance, in an attempt to


6

understand and model pit corrosion (both the behavior of a single corrosion pit, and the probabilistic properties of corrosion pits along a pipeline) on ductile iron pipe, Rajani et al. found:

“No single soil property or combination thereof appeared to emerge as statistically significant in the three data sets (Calgary, Kansas City, Louisville). … It can be concluded that the data on corrosion pits and soil properties from three different cities did not provide any compelling evidence to suggest that the knowledge of soil properties along the pipe improved the ability to predict corrosion pit properties in any significant or consistent manner. … This apparent lack of impact of soil properties on corrosion pit geometry seems to contradict a large body of work in the literature. … There are some indications in the literature that Linear Polarization Resistance (LPR) of soil may be a useful predictor of corrosion in buried pipes.” (Rajani et al. 2011)

According to the ASTM standard addressing polarization resistance measurements in an

aqueous fluid, “Polarization resistance can be related to the rate of general corrosion for metals at or near their corrosion potential, Ecorr. Polarization resistance measurements are an accurate and rapid way to measure the general corrosion rate (ASTM 2014).” In addition, “In this test method, a small potential scan…is applied to a metal sample. The resultant currents are recorded. The polarization resistance, Rp, of a corroding electrode is defined …as the slope of a potential versus current density plot at i=0 (ASTM 2014).” LPR techniques take advantage of the nearly linear relationship between the potential versus current density plot at and near the freely corroding surface potential of that metal. The linear relationship normally extends +10 milliVolts or possibly +20 milliVolts of the freely corroding surface potential depending on the reference consulted (Benjamin et al. 1996, Farrag 2010). The data defining the line in this technique are generated by making short duration polarization offset readings within the linear range. The polarization resistance is inversely related to corrosion rate. The laboratory procedure for estimating corrosion rates from the LPR measurements are described in the ASTM standard (ASTM 2014). In addition, the LPR technique can give a qualitative analysis of the potential for pitting corrosion. Pitting occurs when there is an uneven distribution of anodic and cathodic areas or when these two areas are not in close proximity to each other. This tends to make one electrode more anodic or cathodic than the other. This “current imbalance” between the electrodes can be detected (Rohrback Cosasco Systems, Inc. 2014)

Since this LPR technique includes both the material being corroded and the electrolyte, when applied to soils it provides a comprehensive evaluation of the corrosion potential of that soil system so that it may not be necessary to understand the individual soil and environmental variables, such as chloride, sulfide, etc., that impact corrosion rates. The body of knowledge on application of the LPR technique to soils is still evolving. A study of LPR for the Department of Transportation concluded: “…the laboratory and field measurements of corrosion rates using the LPR device showed uncertainty about their validity in partially and totally dry soils, which makes it difficult to obtain a reliable estimate of the general corrosion rate. Additionally, the average long-term measurement of the LPR in the field did not correlate with the general corrosion rates from weight-loss measurements of buried coupons” (Farrag 2010). However, in some recent work evaluating existing utility soil and LPR data indicates that LPR data reliably correlate to corrosion rates in the tested soils, and that a relationship between maximum pit depth and polarization resistance does exist for pipes of similar age and manufacture method, but the technique is sensitive to the types of metal being evaluated (Dafter 2014).


7

In application to measurement of corrosion of pipelines, LPR has primarily been used in Australia. This technology had been commercially developed and offered by a sole supplier, originally Tyco Pipeline Condition Assessment or Tyco PCA, although Tyco PCA has further evolved in association with Echologics out of Canada. Tyco PCA had developed the proprietary Weibull-based algorithms necessary to translate the raw LPR corrosion rate data into useful information for pipeline management of centrifugally cast gray iron pipe. For the remainder of this report this type of pipe will simply be referred to as cast iron (CI) pipe. The LPR technique has been used by Sydney water and other Australian utilities, who have found the technique to be useful in predicting areas of high CI corrosion and scheduling further evaluation of the pipe (Dafter 2014).

The existing LPR technique involves collection of soils from locations as close as possible to the pipeline to be assessed, both in terms of horizontal and vertical proximity. These soils are transported to the laboratory where they are carefully assessed and prepared, placed in a soil box for testing, and then LPR measurements made at a given soil moisture content which is understood to be the “wilt point” soil moisture level. While non-destructive, and applicable to pipelines still in service, the removal of the soils from the ground and transfer to the laboratory to generate LPR data results in cost and delay in data generation.

There are also methods to assess soils for corrosion potential in order to improve the design of new pipelines. AWWA Standard C-105 includes such a technique that is used to determine whether ductile iron pipe to be buried in a given soil requires additional external corrosion protection. This method includes the determination of several soil parameters, such as pH, resistivity, presence of sulfides, oxidation-reduction potential, and presence of stray currents. These variables are inputs to a calculation set forth in the standard providing a point score. Scores above 10 are considered indicative of corrosive sols and ductile iron pipes in these soils require additional corrosion protection in the form of polyethylene encasement (AWWA 2010). Similarly, the US Bureau of Reclamation has recommended approaches for assessing soil conditions and based on these data providing appropriate corrosion for a new pipeline. However, these techniques guide design and construction of new pipeline installations; they are not intended for evaluation of the condition of existing pipelines.


9

CHAPTER 3 LABORATORY TESTING PLAN

Prior to initiating laboratory testing, changes were made in the original LPR corrosion rate

probe. This probe had been created for an earlier project but improvements to the probe were envisioned to make it more field-useful and more accurate for measurement of ductile iron corrosion. The original LPR probe is shown in Figure 2, below. First, the probe was reduced in size from 1.5-inch diameter down to 1-inch diameter. It was felt that this would make it easier to use in the field. A smaller diameter rod creates a pilot hole where the LPR probe is then inserted. The pilot hole assures maximum contact with the soil without having to work the rod to create a hole. Secondly, and most importantly, the ductile iron rings were subjected to a heat-treating cycle to create an annealing oxide on their exterior surface. It was felt that this would be more representative of actual ductile iron pipe in service. Lastly, the cable connections were brought out through the side of the handles rather than through the top. It was felt that the wires were less likely to be compromised or damaged in this new location. The prototype field LPR probe is presented in Figure 3, below. The probe origins and modifications are presented in Appendix C. Two probes with annealing oxide ductile iron rings were fabricated.

Figure 2 Original LPR probe with ductile iron rings


10

Figure 3 Final prototype field LPR probe with DI rings

The new LPR corrosion rate probes were used in conjunction with the AquaMate™ Portable CORRATER® Instrument. This pairing of the probe and meter was based on the expectation that these devices would likely work together, and the Aquamate CORRATER is the only device known to the researchers for this application. Thus, the data generated by the prototype probe were being read by the Aquamate CORRATER. At the outset of this work, it was not known if the Aquamate CORRATER would work with the probe, nor how reproducible the generated data would be. The typical LPR soil measurements are made in a lab under controlled conditions, especially as regards soil moisture content of the tested soil samples. While the soil moisture can be controlled in the laboratory measurement of LPR, it cannot be controlled in the field; the probe must be capable of data generation under widely varying soil moisture conditions. The literature included examples of LPR data being uncertain in partially or totally dry soils (Farrag 2010).

Soil samples were collected by 8 water utilities (24 samples) in their service area, as well as, collected by Corrpro personnel during external corrosion direct assessment (ECDA) digs along pipelines at 2 natural gas distribution utilities (9 samples) and at one product transport company (3 samples). It should be noted that one additional sample (Sample 32) was received but not tested since it was contaminated during shipping. This is why the tabulated data has a last data point of “Sample 37” but there are only 36 points in actuality.

Soil samples were received in the Corrpro’s Houston Soils Lab and stored until LPR probe testing could be conducted at the same time as laboratory testing. Buckets were received during the spring and summer of 2013 and tested during the winter. Buckets were sealed with tape and stored just outside of the Lab in the warehouse. The revised test plan, presented below as Table 2, was drafted to show how soils would be tested upon receipt at the Soils Lab.


11

Table 2 Soils Testing Plan

As shown in Table 2 there were three stages of testing. First, there was “Bucket Testing” which was intended to simulate field testing of the LPR probes paired with Aquamate CORRATER instruments. These previously untested prototype field LPR probes were tested in uncontrolled conditions to generate LPR corrosion rate data, current imbalance (corrosion pitting index or pitting potential) data, and soil conductivity data. These data generated under these uncontrolled “field” conditions needed to be shown to be consistent and reproducible. It was necessary to verify reproducibility between readings with one prototype LPR probe and then demonstrate relative agreement between the two prototype LPR probes. These readings were intended to simulate field conditions for probe use. Multiple readings were taken in the same bucket changing AquaMate meters each time to verify measurement reproducibility. Readings were also taken over increasing time periods to verify what period of time was necessary for the probe to stabilize in its readings. Lastly, probes were placed side by side and measurements were performed as shown in Figure 4. This was done for each soil sample. Readings were taken on each probe using separate AquaMate meters. This provided the LPR corrosion rate, the imbalance and conductivity specified under Bucket Testing. These measurements verified that the prototype probes created for this project generated reproducible data from the same probe, using the same


12

hole and “sampling” procedures, and between probes and meters. For the prototype field probes to be successful they needed to be stable enough to generate reproducible data, and not be too sensitive to procedural issues such as multiple re-insertions into the soil, and generate similar data between multiple probes.

Figure 4 Side-by-side testing of LPR probes Second, there was “Lab Testing” of soils. These Lab Data, generated in a more traditional

soil box testing apparatus were needed for comparison with the data from the prototype field LPR probes. As each bucket of soil was tested with the LPR probes, a sample of soil was also taken for Lab Testing. The coordination of the Bucket Testing and Lab Testing was especially important since soil moisture levels are known to have a significant impact on resistivity and LPR measurements, and both sets of data represent soil field conditions. The soil box data were generated essentially following ASTM Standard G187-05, Standard Method for Measurement of Soil Resistivity Using the Two-Electrode Soil Box Method in terms of the construction and geometry of the soil box (ASTM 2005). Thick DI plates (full wall ductile iron pipe pieces with a peened surface and annealing oxide) were used as the electrodes. The plates are shown below in Figure 5 and then assembled in the two-electrode soil box (without soil) in Figure 6. Data were generated for LPR corrosion rate, current imbalance, and conductivity. In addition, data were also generated for soil resistivity using a soil box and thick DI plates paired with a Nilsson Soil Resistivity Meter Model 400, at the same soil field moisture conditions.


13

Figure 5 DI thick plates

Figure 6 DI thick plates assembled in two-electrode soil box without soil Third, there was “Standard Laboratory Testing” where testing was done according to

standard approaches, with the exception of the LPR data which have no established standards. However, to better correlate with standard resistivity measurements made at saturated soil conditions, the LPR data were also generated at saturated soil conditions. Thus, a third set of LPR data, current imbalance, and soil conductivity data were taken using the thick DI plate electrode


14

soil boxes, but after drying and saturating the soils. Soil resistivity measurements were also made with a Nilsson Soil Resistivity Meter Model 400, under standard saturated soil conditions (ASTM 2005). Other soil parameters measured that were used to help characterize the corrosive properties of the soil samples included: included: moisture content (by ASTM D2216, Standard Test Methods for Laboratory Determination of Water (Moisture) Content of Soil and Rock by Mass); pH (ASTM G51, Standard Test Method for Measuring pH of Soil for Use in Corrosion Testing), chloride content, and sulfide content. The three sets of data should all correlate for the field probes to be considered successful.

The AquaMate solution conductivity data was converted to resistivity simply based on the fact that conductivity (µmhos/cm = micro mhos per centimeter or µS/cm = micro Siemen per centimeter) is the reciprocal of resistivity (Ω*cm = Ohm centimeter). Using Equation 1.0, below:

As mentioned above, soil resistivity was measured in the two-electrode soil box using both

the Nilsson Soil Resistance Meter per ASTM G187 and the AquaMate Meter. The equipment setups without soil in the soil box are presented in Figures 7 and 8, respectively.

Figure 7 Nilsson Soil Resistance Meter connected to two-electrode soil box without soil

1=Resistivity

AquaMateDisplaySolutionConductivity

(µS/cm) Eq1.0

CalculatedResistivity (Ω*cm)

1=SolutionConductivity


15

Figure 8 AquaMate Meter connected to two-electrode soil box without soil


17

CHAPTER 4 LABORATORY TESTING RESULTS AND DISCUSSION

Considerable data were generated in this study. Raw data are presented in Appendix D.

There is no standard addressing the generation of LPR data, and so American Society of Testing and Materials (ASTM) Standard G187-05, “Standard Test Method for Measurement of Soil Resistivity Using the Two-Electrode Soil Box Method,” guided the generation and interpretation of many of the data in this study (ASTM 2005). LPR data were correlated with resistivity data through all conditions studied. For resistivity data ASTM G187 Section 13, Precision and Bias, indicates that the 95% confidence interval for Repeatability is 18.5% and for Reproducibility is 29.7% of average readings. ASTM G187 addresses soil resistivity measurements done in a laboratory under controlled and standard conditions. Greater variations might be allowed in field measurements due to widely variable field conditions, as was simulated in this study with bucket testing, but 18.5% was used as the criterion for repeatable measurements.

Initial Bucket Testing of the LPR corrosion rate probes was performed to establish measurement reproducibility. Multiple readings were taken in the same bucket changing AquaMate meters each time. It was determined that the meter did not affect the readings. Further, the probe was extracted and reinserted into the soil several times for additional reading. While not exactly the same, the values were within 18.5% of each other.

Readings were also taken over increasing time periods to verify how long probe-soil contact time needs to be for the probe readings to stabilize. There were also concerns that probe readings might fluctuate over time. This was evaluated by taking multiple readings over increasing time periods of 30 seconds, 2 minutes, and 10 minutes. Table 3 presents the corrosion rate data measured over time. These simulated field tests indicate that probe readings have stabilized after 2 minutes of contact time with the soil. Probe data reported in Tables 4 onward are two-minute data. However, field-testing should be conducted before a final timeframe is determined for probe stabilization. Field conditions will affect the probe readings and might result in the need for longer probe-soil contact times to achieve stable readings.

Data also had to be generated to determine if two probes placed into the same soil will generate the same data. The probes were placed side by side in the soil sample buckets and measurements were performed (Table 4 and Graph 1). While exactly similar readings were not expected, extremely close reading were noted in most instances, and the data were judged close enough to have confidence to move forward with additional testing. Differences varied from 4% to 24% or an average of 14%, which is acceptable for readings in soil based on ASTM G 187 Section 13, as discussed above. Readings were probably affected by surface contact with the surrounding soil, and variation in the annealing oxide on the LPR ductile iron rings. While the annealing oxide gives the rings similar properties to pipe in the ground, the oxide can be variable from ring to ring and probe to probe and it likely affects the LPR corrosion rate readings. The exact differences in the annealing oxide on the rings was not quantified or tested, but developing a standard method for doing this should be considered in future research.


18

Table 3 Bucket testing corrosion rate data over time

Bucket Testing

Sample Number Time

Probe 1 Corrosion

Rate (mpy)

Probe 2 Corrosion

Rate (mpy)

10 2 min 0.30 0.27 10 10 min 0.28 0.27 11 2 min 0.03 0.11 11 10 min 0.03 0.13 12 2 min 0.22 0.32 12 10 min 0.23 0.33 19 2 min 0.39 0.61 19 10 min 0.44 0.61 20 2 min 0.04 0.18 20 10 min 0.05 0.19 35 2 min 0.78 0.73 35 10 min 1.25 0.79


19

Table 4 Corrosion rate data (two-minute probe readings)

Bucket Testing Lab

Testing

Sample Number

Probe 1 Corrosion

Rate (mpy)

Probe 2 Corrosion

Rate (mpy)

Probe Average

CR (mpy)

DI Thick Plates CR

(mpy) 1 0.76 1.77 1.27 0.87 2 0.37 0.39 0.38 0.42 3 1.02 1.29 1.16 0.65 4 0.63 0.33 0.48 0.86 5 1.37 1.42 1.40 1.89 6 2.30 3.14 2.72 3.58 7 0.52 0.80 0.66 2.29 8 0.16 0.11 0.14 0.43 9 0.88 0.44 0.66 1.84

10 0.28 0.27 0.28 0.98 11 0.08 0.04 0.06 0.31 12 0.23 0.33 0.28 1.09 13 0.34 0.10 0.22 0.14 14 0.21 0.19 0.20 0.12 15 0.62 0.56 0.59 0.15 16 1.38 0.99 1.19 0.76 17 1.83 1.92 1.88 4.95 18 2.42 1.90 2.16 2.69 19 1.14 1.58 1.36 3.10 20 0.05 0.19 0.12 0.47 21 0.25 0.11 0.18 0.14 22 1.58 2.43 2.01 9.91 23 2.06 3.12 2.59 8.89 24 1.60 1.22 1.41 0.73 25 1.87 2.13 2.00 3.28 26 1.49 2.20 1.85 2.63 27 1.73 1.98 1.86 5.68 28 3.09 2.49 2.79 5.59 29 1.75 2.25 2.00 2.44 30 2.31 2.46 2.39 12.3 31 0.32 0.26 0.29 0.02 33 0.19 0.08 0.14 0.27 34 0.38 0.26 0.32 0.06 35 2.24 1.86 2.05 3.03 36 0.22 0.21 0.22 0.31 37 0.15 0.07 0.11 0.19


20

Graph 1, below, presents the readings after 2 minutes from each probe taken simultaneously in the soil sample buckets.

Graph 1

The corrosion rate data from side-by-side Bucket Testing was compared to the Lab Testing measured corrosion rate in a two-electrode soil box with thick DI plates and same soil moisture conditions. These data are presented in Table 4 and are plotted in Graph 2. The data within each technique track well but the data from the two-electrode soil box had higher corrosion rates than the LPR corrosion rate probes. There are not enough data to determine the exact ratio between the two approaches. Several extreme outlier points skew the graph and possible multiplier. More data in the future may reduce the impact of the outliers and show a tighter correlation.

Imbalance or corrosion pitting index, a unit less value, was also analyzed between the two LPR probes. When the imbalance reading is less than the corrosion rate reading or close to zero, corrosion is uniform corrosion with insignificant pitting. If the imbalance is greater than the corrosion rate value, this is indicative of increased pitting. Table 5 presents the average CR and average imbalance for the corrosion rate probes. As shown in Graph 3, as the CR increased, the imbalance increased. This agrees with the understanding of the corrosion process that a higher corrosion rate is indicative of pitting corrosion rather than general or uniform corrosion. Graph 4 combines the data into average CR and imbalance for the probes. Table 5 also presents the same information for the two-electrode soil box with thick DI plates. Graph 5 plots the same data for two-electrode soil box using the thick ductile iron plates. While the readings between the Bucket Testing probe data and the Lab Testing data are not the same, they do demonstrate the same trend; higher corrosion rate correlates to a higher imbalance or pitting potential.


21

Graph 2

Table 6 presents the LPR and resistivity data generated in this project from the LPR probe, used under simulated field conditions (“Bucket Testing” data), and the data generated by more standard approaches, although under non-standard conditions, primarily field soil moisture conditions (“Lab Testing” data). In each case, both LPR (corrosion rate) and resistivity data are being compared. Resistivity was calculated from the Aquamate CORRATER conductivity data using Equation 1.0. These data were generated according to the plan laid out in Table 2, previously discussed.

Graphs 6 and 7 compare the LPR probe-generated data with the lab-generated date. In each case, the LPR (corrosion rate) and resistivity data are compared since it is generally accepted that there is a strong correlation between soil resistivity and corrosion, namely, less resistive soils are more corrosive. This relationship is seen in Graph 6 where each probe has essentially the same response to test conditions, and corrosion rate increases with decreasing soil resistivity. The same relationship can be noted in Graph 7 with the data generated from the two-electrode soil box lab testing. Resistivity calculated from the Aquamate CORRATER data and the resistivity data take directly from the Nilsson Meter, are in close correlation, nearly identical at the noted scale. However, at any resistivity below approximately 10,000 ohm-cm, the corrosion rate is lower in the LPR probe generated data as compared with the soil box data. While these differences can be noted, they cannot be resolved as to which, if either, of the two methods of corrosion and resistivity data generation might be more accurate, since the project team lacks the field data for correlation of corrosion losses, especially pitting, with measurements of corrosion rate or resistivity.

Standard Laboratory Tests data is summarized in Table 7, below. The table presents: moisture content; pH; chloride content; sulfide content; conductivity using conductivity probe; calculated resistivity from conductivity data; and resistivity (on dried then saturated samples) calculated from AquaMate conductivity data and measured with the Nilsson meter. Soil classification for each sample was based on the standard calculated resistivity value. Using Table 1 from USDOT Publication No. FHWA-NH1-00-44, Corrosion Degradation of Soil


22

Reinforcements for Mechanically Stabilized Earth Walls and Reinforced Soil Slopes, and presented below as Table 8, classifications were assigned for each soil sample.

Table 5 Corrosion rate and imbalance data

Bucket Testing Lab Testing

Sample Number

Probe Average

CR (mpy)

Probe Average

Imbalance

DI Thick Plates CR

(mpy)

DI Thick Plates

Imbalance 1 1.27 6.51 0.87 2.85 2 0.38 1.69 0.42 0.77 3 1.16 5.32 0.65 1.29 4 0.48 2.61 0.86 6.59 5 1.40 7.28 1.89 0.02 6 2.72 12.95 3.58 0.49 7 0.66 3.28 2.29 0.39 8 0.14 0.28 0.43 0.19 9 0.66 4.26 1.84 2.13

10 0.28 1.78 0.98 1.02 11 0.06 0.51 0.31 0.77 12 0.28 1.84 1.09 1.35 13 0.22 1.09 0.14 0.03 14 0.20 0.33 0.12 0.23 15 0.59 5.24 0.15 0.82 16 1.19 8.59 0.76 0.16 17 1.88 8.10 4.95 6.39 18 2.16 11.10 2.69 4.82 19 1.36 8.37 3.10 1.40 20 0.12 1.05 0.47 1.08 21 0.18 0.83 0.14 1.52 22 2.01 11.17 9.91 10.40 23 2.59 12.00 8.89 6.31 24 1.41 4.23 0.73 1.78 25 2.00 7.88 3.28 5.94 26 1.85 8.39 2.63 1.91 27 1.86 10.02 5.68 2.80 28 2.79 11.30 5.59 8.72 29 2.00 13.60 2.44 4.39 30 2.39 11.24 12.3 11.90 31 0.29 0.64 0.02 0.25 33 0.14 0.15 0.27 0.72 34 0.32 0.61 0.06 0.15 35 2.05 26.55 3.03 9.10 36 0.22 0.27 0.31 1.43 37 0.11 0.22 0.19 0.29


23

Graph 3

Graph 4


24

Graph 5


25

Table 6 Corrosion rate and resistivity data

Sample Number

Probe 1 Corrosion

Rate(mpy)

Probe 1 Resistivity AquaMate Calculated(ohm-cm)

Probe 2 Corrosion

Rate(mpy)


Probe Average

ResistivityAquaMate Calculated(ohm-cm)

DI Thick Plates CR

(mpy)

DI Thick Plates

Resistivity AquaMate Calculated(ohm-cm)

DI Thick Plates

Resistivity Nilsson

(ohm-cm)

1 0.76 1,969 1.77 2,212 2,090 0.87 3,040 3,2002 0.37 11,236 0.39 18,519 14,877 0.42 10,753 11,0003 1.02 4,630 1.29 5,319 4,974 0.65 6,410 6,0004 0.63 5,618 0.33 7,143 6,380 0.86 3,704 3,6005 1.37 1,905 1.42 3,040 2,472 1.89 3,610 3,4006 2.30 1,183 3.14 613 898 3.58 1,377 1,2007 0.52 11,628 0.80 12,987 12,307 2.29 2,451 2,3008 0.16 33,333 0.11 76,923 55,128 0.43 10,309 10,0009 0.88 6,667 0.44 10,000 8,333 1.84 2,660 2,50010 0.28 4,292 0.27 6,494 5,393 0.98 9,524 8,80011 0.08 34,483 0.04 125,000 79,741 0.31 18,182 16,00012 0.23 11,628 0.33 7,692 9,660 1.09 6,849 5,60013 0.34 24,390 0.10 100,000 62,195 0.14 90,909 84,00014 0.21 200,000 0.19 200,000 200,000 0.12 45,455 52,00015 0.62 2,364 0.56 2,545 2,454 0.15 5,435 6,80016 1.38 4,405 0.99 5,780 5,093 0.76 5,128 9,60017 1.83 3,257 1.92 2,890 3,074 4.95 3,690 3,60018 2.42 437 1.90 887 662 2.69 883 48019 1.14 2,237 1.58 2,710 2,474 3.10 3,448 2,90020 0.05 142,857 0.19 32,258 87,558 0.47 27,027 13,00021 0.25 23,256 0.11 111,111 67,183 0.14 7,246 7,60022 1.58 1,238 2.43 1,056 1,147 9.91 1,805 1,30023 2.06 3,030 3.12 2,985 3,008 8.89 3,378 2,70024 1.60 5,882 1.22 3,497 4,689 0.73 4,902 3,10025 1.87 4,405 2.13 3,333 3,869 3.28 3,185 2,40026 1.49 6,024 2.20 6,579 6,302 2.63 9,346 36,00027 1.73 3,311 1.98 4,049 3,680 5.68 5,525 6,00028 3.09 812 2.49 833 823 5.59 1,193 1,10029 1.75 1,548 2.25 1,645 1,596 2.44 2,304 1,70030 2.31 556 2.46 576 566 12.3 832 80031 0.32 32,258 0.26 45,455 38,856 0.02 200,000 18,00033 0.19 83,333 0.08 500,000 291,667 0.27 52,632 36,00034 0.38 50,000 0.26 40,000 45,000 0.06 90,909 72,00035 2.24 790 1.86 1,984 1,387 3.03 1,342 1,10036 0.22 142,857 0.21 333,333 238,095 0.31 43,478 72,00037 0.15 500,000 0.07 500,000 500,000 0.19 52,632 56,000



26

Graph 6

Graph 7


27

Table 7 Standard laboratory testing data

Sample Number

Moisture Content

% pHChlorides

(ppm)Sulfides (ppm)

Standard Electrode

Conductivity (µmhos/cm)

Standard Calculated Resistivity (ohm-cm) Sample Type

Soil Classification*

DI Thick Plates CR

(mpy)

DI Thick Plates

Imbalance

DI Thick Plates


DI Thick Plates

Resistivity Nilsson

(ohm-cm)

1 28.0 8.3 4 0 610 1,600 Clay loam Corrosive 1.34 4.38 2,110 2,1002 16.0 7.0 2 0 170 5,900 Clay loam Mildly Corrosive 1.41 0.40 5,848 5,2003 28.0 6.9 4 0 450 2,200 Sandy loam & clay Moderately Corrosive 1.26 0.77 2,994 2,7004 10.0 7.9 60 0 1,000 1,000 Silty clay Corrosive 3.18 2.80 2,381 2,0005 25.0 7.7 42 0 390 2,600 Silty clay Moderately Corrosive 2.06 1.49 3,049 2,5006 17.0 7.8 160 0 1,500 670 Sandy clay Very Corrosive 3.65 1.33 1,350 1,2007 14.0 7.9 420 0 2,300 440 Silty loam & rocks Very Corrosive 2.64 1.41 1,048 1,0008 10.0 7.7 20 0 670 1,500 Sandy clay Corrosive 1.47 0.35 3,096 2,9009 19.0 7.3 84 0 1,000 1,000 Sandy clay Corrosive 3.06 3.02 1,976 1,700

10 23.0 7.3 28 0 400 2,500 Sandy clay loam & rocks Moderately Corrosive 1.44 0.11 5,263 5,200

11 17.0 7.2 2 0 120 8,300 Sandy clay loam Mildly Corrosive 1.04 0.53 8,696 7,600

12 22.0 7.9 33 0 530 1,900 Sandy clay loam & rocks Corrosive 1.95 0.93 2,825 2,500

13 3.8 6.9 1 0 120 8,300 Fine sand Mildly Corrosive 0.93 0.99 18,868 16,000

14 5.3 7.7 4 0 380 2,600 Silty loam & rocks Moderately Corrosive 1.46 0.05 4,608 4,400

15 16.0 6.3 4 0 600 1,700 Sandy clay loam Corrosive 1.43 5.13 1,783 1,600

16 19.0 6.3 7 0 310 3,200 Sandy clay Moderately Corrosive 3.45 4.29 3,497 3,000

17 34.0 6.3 6 0 660 1,500 Silty clay Corrosive 6.95 6.26 2,551 2,700

18 18.0 7.4 420 0 3,000 330 Sandy clay Very Corrosive 4.13 3.73 744 380

19 36.0 7.0 15 0 300 3,300 Rocks & clay loam Moderately Corrosive 2.71 2.59 3,367 2,900

20 15.0 4.7 14 0 150 6,700 Rocks & clay loam Mildly Corrosive 1.10 0.51 10,309 8,000

21 11.0 8.1 10 0 220 4,500 Rocks & clay loam Moderately Corrosive 1.61 1.74 4,082 3,200

22 28.0 7.1 4 0 4,100 240 Sandy clay loam Very Corrosive 14.6 10.9 1,049 1,000

23 20.0 7.7 96 0 650 1,500 Sandy loam Corrosive 8.75 7.94 2,404 2,000

24 25.0 5.0 160 0 820 1,200 Silty clay loam Corrosive 2.78 2.08 1,639 1,400

25 45.0 7.7 33 0 480 2,100 Silty loam & rocks Moderately Corrosive 3.26 5.96 3,040 2,400

26 43.0 4.6 10 0 110 9,100 Silty clay loam & rocks Mildly Corrosive 2.73 1.23 9,091 36,000

27 19.0 4.6 8 0 410 2,400 Clay loam Moderately Corrosive 5.88 2.83 5,495 5,600

28 39.0 5.1 290 0 1,400 710 Clay Corrosive 5.99 10.3 1,192 920

29 19.0 7.4 84 0.4 700 1,400 Clay Corrosive 2.85 3.76 2,299 1,700

30 21.0 7.6 250 0 1,400 710 Clay Corrosive 11.7 9.46 835 800

31 10.0 7.1 3 0 150 6,700 Sandy clay loam Mildly Corrosive 1.15 1.45 9,524 7,600

33 16.0 6.6 12 0 160 6,300 Sand & clay Mildly Corrosive 3.12 2.04 12,658 14,000

34 8.0 7.7 1 0 100 10,000 Sand & rocks Mildly Corrosive 0.83 0.71 23,256 16,000

35 19.0 4.4 6 0 1,700 590 Clay Very Corrosive 3.16 5.55 1,495 960

36 11.0 5.3 1 0 49 20,000 Sandy clay loam Non-Corrosive 2.55 2.10 17,241 15,000

37 10.0 5.0 1 0 33 30,000 Sandy clay loam Non-Corrosive 0.54 0.37 26,316 19,000

*FHWA-NH1-00-44, Corrosion Degradation of Soil Reinforcements for Mechanically Stabilized Earth Walls and Reinforced Soil Slopes

Saturated Sample

Standard Laboratory Testing


28

Table 8 Effect of resistivity on corrosion

Aggressiveness Resistivity (ohm-cm) Very Corrosive < 700

Corrosive 700 - 2,000 Moderately Corrosive 2,000 - 5,000

Mildly Corrosive 5,000 - 10,000

Non-Corrosive > 10,000 FHWA-NH1-00-44, Corrosion Degradation of Soil Reinforcements for Mechanically Stabilized Earth Walls and Reinforced Soil Slopes

Other Standard Laboratory Test data that was compared to the corrosion rate included soil

sample chloride content and pH data, which are presented in Table 8. The chloride content for each soil sample was plotted against the average corrosion rates of the LPR probes, against the corrosion rates of the thick ductile iron plates in the soil box and against the corrosion rate of the thick ductile iron plates in the soil box after the soil was dried then saturated. These plots are presented in Graphs 8, 9 and 10, respectively. The data with the best correlation is the corrosion rate from the Bucket Testing. It was interesting to note that the pH value was not viewed as a significant contributing factor to the soils’ corrosion rate. This is demonstrated in Graphs 11 and 12.


29

Table 9 Corrosion rate, pH and chloride data

Bucket Testing

Lab Testing

Sample Number

Probe Average CR

(mpy)

DI Thick Plates CR

(mpy) pHChlorides

(ppm)

1 1.27 0.87 8.3 42 0.38 0.42 7.0 23 1.16 0.65 6.9 44 0.48 0.86 7.9 605 1.40 1.89 7.7 426 2.72 3.58 7.8 1607 0.66 2.29 7.9 4208 0.14 0.43 7.7 209 0.66 1.84 7.3 84

10 0.28 0.98 7.3 28

11 0.06 0.31 7.2 2

12 0.28 1.09 7.9 33

13 0.22 0.14 6.9 1

14 0.20 0.12 7.7 4

15 0.59 0.15 6.3 4

16 1.19 0.76 6.3 7

17 1.88 4.95 6.3 6

18 2.16 2.69 7.4 420

19 1.36 3.10 7.0 15

20 0.12 0.47 4.7 14

21 0.18 0.14 8.1 10

22 2.01 9.91 7.1 4

23 2.59 8.89 7.7 96

24 1.41 0.73 5.0 160

25 2.00 3.28 7.7 33

26 1.85 2.63 4.6 10

27 1.86 5.68 4.6 8

28 2.79 5.59 5.1 290

29 2.00 2.44 7.4 84

30 2.39 12.3 7.6 250

31 0.29 0.02 7.1 3

33 0.14 0.27 6.6 12

34 0.32 0.06 7.7 1

35 2.05 3.03 4.4 6

36 0.22 0.31 5.3 1

37 0.11 0.19 5.0 1



30

Graph 8

Graph 9


31

Graph 10

Graph 11


32

Graph 12

While it is understood that the soil used in the Standard Laboratory Tests has been dried

and then saturated prior to testing, it was decided to plot this data against the data obtained for the simultaneous Bucket Testing and Lab Testing. For ease of comparison, the data is presented in Tables 10 and 11. The data for both corrosion rate and resistivity track closely to the data obtained during earlier testing. As shown in Graph 13, the corrosion rate on the soil samples that were dried and then saturated seem to run higher than the CR obtained during Bucket Testing and Lab Testing. The resistivity measured by various techniques is presented in Graph 14. As a reflection that the corrosion rate runs higher in the Standard Laboratory Tests, the calculated and measured resistivities are lower than the earlier obtained Bucket Testing and Lab Testing data.


33

Table 10 Comparison of corrosion rate data

Bucket Testing

Lab Testing

Standard

Laboratory

Testing

Sample Number

Probe Average

CR (mpy)

DI Thick Plates

CR (mpy)

Saturated DI Thick

Plates CR (mpy)

1 1.27 0.87 1.342 0.38 0.42 1.413 1.16 0.65 1.264 0.48 0.86 3.185 1.40 1.89 2.066 2.72 3.58 3.657 0.66 2.29 2.648 0.14 0.43 1.479 0.66 1.84 3.06

10 0.28 0.98 1.44

11 0.06 0.31 1.04

12 0.28 1.09 1.95

13 0.22 0.14 0.93

14 0.20 0.12 1.46

15 0.59 0.15 1.43

16 1.19 0.76 3.45

17 1.88 4.95 6.95

18 2.16 2.69 4.13

19 1.36 3.10 2.71

20 0.12 0.47 1.10

21 0.18 0.14 1.61

22 2.01 9.91 14.6

23 2.59 8.89 8.75

24 1.41 0.73 2.78

25 2.00 3.28 3.26

26 1.85 2.63 2.73

27 1.86 5.68 5.88

28 2.79 5.59 5.99

29 2.00 2.44 2.85

30 2.39 12.3 11.7

31 0.29 0.02 1.15

33 0.14 0.27 3.12

34 0.32 0.06 0.83

35 2.05 3.03 3.16

36 0.22 0.31 2.55

37 0.11 0.19 0.54


34

Table 11 Comparison of resistivity data

Bucket Testing

Sample Number

Probe Average


DI Thick Plates


DI Thick Plates

Resistivity Nilsson

(ohm-cm)

Saturated DI Thick

Plates Resistivity AquaMate Calculated(ohm-cm)

SaturatedDI Thick

Plates Resistivity

Nilsson (ohm-cm)

1 2,090 3,040 3,200 2,110 2,1002 14,877 10,753 11,000 5,848 5,2003 4,974 6,410 6,000 2,994 2,7004 6,380 3,704 3,600 2,381 2,0005 2,472 3,610 3,400 3,049 2,5006 898 1,377 1,200 1,350 1,2007 12,307 2,451 2,300 1,048 1,0008 55,128 10,309 10,000 3,096 2,9009 8,333 2,660 2,500 1,976 1,700

10 5,393 9,524 8,800 5,263 5,200

11 79,741 18,182 16,000 8,696 7,600

12 9,660 6,849 5,600 2,825 2,500

13 62,195 90,909 84,000 18,868 16,000

14 200,000 45,455 52,000 4,608 4,400

15 2,454 5,435 6,800 1,783 1,600

16 5,093 5,128 9,600 3,497 3,000

17 3,074 3,690 3,600 2,551 2,700

18 662 883 480 744 380

19 2,474 3,448 2,900 3,367 2,900

20 87,558 27,027 13,000 10,309 8,000

21 67,183 7,246 7,600 4,082 3,200

22 1,147 1,805 1,300 1,049 1,000

23 3,008 3,378 2,700 2,404 2,000

24 4,689 4,902 3,100 1,639 1,400

25 3,869 3,185 2,400 3,040 2,400

26 6,302 9,346 36,000 9,091 36,000

27 3,680 5,525 6,000 5,495 5,600

28 823 1,193 1,100 1,192 920

29 1,596 2,304 1,700 2,299 1,700

30 566 832 800 835 800

31 38,856 200,000 18,000 9,524 7,600

33 291,667 52,632 36,000 12,658 14,000

34 45,000 90,909 72,000 23,256 16,000

35 1,387 1,342 1,100 1,495 960

36 238,095 43,478 72,000 17,241 15,000

37 500,000 52,632 56,000 26,316 19,000

Lab Testing

Standard Laboratory

Testing


35

Graph 13

Graph 14


37

CHAPTER 5 CONCLUSIONS

The strong and enthusiastic response from utilities contacted for this project indicates that

the water and wastewater industry is extremely interested in the issue of determining the current condition and projected life of their buried assets. LPR has the potential to provide a non-destructive assessment of assets that are still in service. LPR is a technique that provides an indirect measurement of the corrosion rate of materials in a given environment. Both Sydney Water and Hunter Water use LPR as a means of screening centrifugally cast gray iron pipelines for areas of higher concern. As currently practiced in Australia, LPR involves field collection of soil samples and laboratory analysis of those samples. In Australia the LPR technique has been commercialized by one organization that has developed algorithms to use the corrosion rate and soils data to project expected lifespan of the given gray iron pipelines. This project focused on development of a field-based LPR probe allowing data collection in the field, and focused on assessment of the corrosion rate of bare ductile iron pipelines.

Primary accomplishments of this project included:

1. Fabricated two examples of a prototype field LPR probe for ductile iron pipe. Two field LPR probes were fabricated in this work, and these probes included annealed ductile iron electrodes to customize these probes for the testing of ductile iron pipe corrosion.

2. Tested different examples of the same commercially available corrosion rate-monitoring meters (AquaMate CORRATER meters) with the same probe to generate LPR data. These different meters were found to generate comparable LPR data on field condition soils when used with the same probe. These results indicate data reproducibility regardless of which meter was used.

3. Tested two field LPR probes in simulated field conditions to generate LPR data on the same soil. These different probes were found to generate comparable LPR and conductivity data on field condition soils for a wide variety of soils. These results indicate data reproducibility regardless of which of the two probes was used.

4. Determined that the field probe readings stabilized quickly once in contact with the test soils. Stable readings could be taken after two minutes of probe-soil contact time under these test conditions.

5. Tested the field LPR probes in 36 soils for the generation of LPR and other data, in particular resistivity/conductivity of the soils. The data generated by the field probes were correlated to other data through a series of tests in non-standard and standard conditions allowing comparison of the different types of data. Results of the testing include:

a. The field probes generated reproducible and consistent LPR measurements when tested on field condition soils. These LPR measurements are an indirect measurement of the corrosion rate between the metal, in this case ductile iron, and the soils in which it is placed.

b. The field probe generated LPR data were compared to LPR data generated by the soil box method under soil field conditions. Good correlation between the two sets of data was found although soil box corrosion rates tended to be higher than the corrosion rates measured by field probes.

c. The field probe generated LPR data were compared to LPR data generated by the soil box method under typical test conditions of saturated soil. No standard exists


38

for generation of LPR data in soils, but saturated soil conditions are a standard test condition set forth for generation of soil resistivity data (ASTM 2005). Good correlation between the two sets of data was found but in nearly all instances, the Standard Laboratory Testing data on saturated soils yielded a higher corrosion rate than the testing of the field-moisture soil samples. The lack of corrosion and pitting data from pipe in the field makes it impossible to determine which corrosion rate correlates better with true field conditions.

d. The field probes generated reproducible and consistent soil conductivity measurements on field condition soils. Soil resistivity is the reciprocal of soil conductivity, and thus can be calculated from soil conductivity.

e. The field probe-generated soil conductivity data were compared to soil conductivity data generated by the soil box method under soil field conditions. Good correlation between the two sets of data was found.

f. The field probe-generated soil conductivity data were compared to soil resistivity data generated by the soil box method under standard saturated soil conditions using the Nilsson resistivity meter (ASTM 2005). Good correlation between the two sets of data was found.

g. The LPR corrosion rate probe and AquaMate CORRATER provide data that trend well with the data obtained using a two-electrode soil box. The conductivity data converted to soil resistivity readings from the LPR probes trended well and had similar values when compared to Laboratory Data and Standard Laboratory Data measured using the AquaMate and Nilsson Soil Resistance Meter. These resistivity values behaved as expected; a decrease in soil resistivity was associated with an increase in corrosion rate.

h. The two-electrode soil box typically yielded higher corrosion rate data when compared with the LPR probes. Direct ratios or multipliers between simulated field condition (Bucket Testing) data and the Lab Testing and Standard Laboratory Testing data have not been calculated at this early juncture based on the small sample population and general lack of field corrosion data. Future research should strive to address these issues by expanding the database.

i. Similarly, measured imbalance values behaved as expected; increasing imbalance values were associated with an increase in corrosion rate. Increased imbalance values can indicate a greater possibility of pitting corrosion.

Unfortunately, the project failed to accomplish the original objective of developing an

empirical model relating soil characteristics to pipe corrosion due to lack of field corrosion data that were available or could be generated for this project. Thirteen of fifteen utilities involved in this project had very limited corrosion data on their ductile iron pipelines. The project also failed to establish a direct correlation between measured corrosion rate and true corrosion loss and/or corrosion pit depths measured on in-service water mains of known age again due to lack of field corrosion data.

The prototype LPR probe seems to be an effective tool to gather data under simulated field conditions that have good confidence when compared to lab-generated data. The ease of use and direct read display of the AquaMate Corrosion Rate Meter makes it a good companion tool to the LPR corrosion rate probe. The relatively low costs of the corrosion rate meter ($1,690.00) and the


39

LPR corrosion rate probe with annealed DI rings ($595.00) would seem to make them affordable for water and wastewater utilities.

Based on the data generated in this project, adequate testing has been completed to verify that a field-based LPR probe appears capable of development. A field-based LPR probe would allow generation of more LPR data more quickly, and these data would be helpful in screening existing ductile iron pipelines for areas of higher corrosion rate. However, to be able to use the LPR corrosion rate probe as a field-ready predictive analysis tool, further studies will be required to establish a direct correlation between measured corrosion rate and actual corrosion losses measured on in-service water mains of known age.


41

CHAPTER 6 RECOMMENDATIONS

The research project has shown considerable promise at rather low cost. Additional

research is required to determine the impact of several factors and more may be identified in the future.

Testing must be done on more probe samples to verify that the tool can be manufactured inexpensively and with reproducibility. The performance characteristics of the annealing oxide on the rings must be evaluated to determine how it will perform or hold up under repeated use in an operating environment known to scratch metal tool surfaces. Questions include:

Is the annealing oxide reproducible with consistency? What happens if it is not exactly the same during manufacturing? Is there some acceptable multiplier to allow the rings use if there is a difference? What is the effect of scratching the oxide? What constitutes a failed tool?

A database must be established to create a large enough soil sample population so that a

relational multiplier can be created and used to directly correlate field data to laboratory analytical results. Additional studies are required that actually fund excavation of pipeline sections of in-service water mains so that direct readings of pit depths can be performed. Once the tool’s corrosion rate is correlated to actual pit depth, it is expected that definite remaining service life of ductile iron pipe can be estimate or predicted.


43

REFERENCES

ASTM (American Society for Testing and Materials). 2005. Standard Test Method for Measurement of Soil Resistivity Using the Two-Electrode Soil Box Method. Designation G187-05. West Coshocken, Pa.: ASTM International.

ASTM (American Society for Testing and Materials). 2014. Standard Test Method for Conducting Potentiodynamic Polarization Resistance Measurements. Designation G59-97. West Coshocken, Pa.: ASTM International.

AWWA (American Water Works Association). 2010. Standard for Polyethylene Encasement for Ductile-Iron Pipe Systems. AWWA C105-10. Denver, Colo.: AWWA.

AWWA (American Water Works Association). 2012. Buried No Longer: Confronting America’s Water Infrastructure Challenge. Denver, Colo.: AWWA. Accessed June 23, 2014. http://www.awwa.org/Portals/0/files/legreg/documents/BuriedNoLonger.pdf

Benjamin, M.M., J.F. Ferguson, O. von Franqué, G.J. Kirmeyer, P. Leroy, R.J. Oliphant, S.H. Reiber, R.A. Ryder, and M.R. Schock. 1996. Internal Corrosion of Water Distribution Systems. Cooperative Research Report, Second Edition. Denver, Colo.: AwwaRF and Karlsruhe, Germany: DVGW-TZW.

Cole, I.S., and D. Marney. 2012. The Science of Pipe Corrosion: A Review of the Literature on the Corrosion of Ferrous Metals in Soil. Corrosion Science, 56: 5 – 16.

Dafter, M. 2014. Using LPR to Predict Underground Corrosion of Cast Iron Water Mains. In 2014 Australian Corrosion & Prevention Conference & Exhibition. September 21 – 24, Darwin Convention Center, Australia.

Farrag, K. 2010. Evaluation of External Corrosion-Rate Using Polarization Resistance and Soil Properties, Final Report. DOT Project No. 256, Contract Number: DTPH56-08-T-000022. Des Plaines, Ill.: Gas Technology Institute. Accessed June 23, 2014. http://ntl.bts.gov/lib/46000/46300/46323/FilGet.pdf

Kirmeyer, G.J., M. Friedman, K. Martel, G. Thompson, A. Sandvig, J. Clement, and M. Frey. 2002. Guidance Manual for Monitoring Distribution System Water Quality. Denver, Colo.: AwwaRF and the American Water Works Association.

Marlow, D., P. Davis, D. Trans, D. Beale, S. Burn, and A. Urquhart. 2009. Remaining Asset Life: A State of the Art Review - Strategic Asset Management. Alexandria, Va.: Water Environment Research Foundation.

NAS (National Academies Press). 2009. Review of the Bureau of Reclamation’s Corrosion Prevention Standards for Ductile Iron Pipe. Washington, D.C.: The National Academies Press. Accessed June 20, 2014. http://www.nap.edu/catalog/12593/review-of-the-bureau-of-reclamations-corrosion-prevention-standards-for-ductile-iron-pipe

Rajani, B., Y. Kleiner, and D. Krys. 2011. Long-Term Performance of Ductile Iron Pipes. Denver, Colo.: Water Research Foundation.

Ricker, R.E. 2010. Analysis of Pipeline Steel Corrosion Data from NBS (NIST) Studies Conducted Between 1922 – 1940 and Relevance to Pipeline Management. Journal of Research of the National Institute of Standards and Technology, 115(5): 373 – 392. http://nvlpubs.nist.gov/nistpubs/jres/115/5/05-j115-5-ricker.pdf

Rohrback Cosasco Systems, Inc. 2014. CORRATER Probe Selection Guide. Bulletin #400-G. Santa Fe Springs, Calif.: Rohrback Cosasco Systems, Inc. Accessed June 23, 2014. http://www.cosasco.com/documents/Datasheets/Corrater_LPR_Probe_Selection_Guide.pdf


44

Romanoff, N. 1957. Underground Corrosion. National Bureau of Standards Circular 579. Washington, D.C.: National Bureau of Standards. Accessed June 20, 2014. https://archive.org/stream/UndergroundCorrosion#page/n3/mode/2up

Romer, A.E., G.E.C. Bell, S.J. Duranceau, and S. Foreman. 2004. External Corrosion and Corrosion Control of Buried Water Mains. Denver, Colo.: AwwaRF.

Silverman, D.C. 1996. Measuring Corrosion Rates in Drinking Water By Linear Polarization – Assumptions and Watchouts. In Proceedings of the 1995 Water Quality Technology Conference. November 12 – 16, 1995, New Orleans, Louisiana.

Vilda, W.S. III; D. Lindemuth; J.A. Ellor; M. Islam; J. Peter Ault; E.C. Flounders, Jr.; and J. Repp. 2009. Corrosion in the Soil Environment: Soil Resistivity and pH Measurements. NCHRP Project 21-06. Washington, D.C.: Transportation Research Board. Accessed June 23, 2014. http://onlinepubs.trb.org/onlinepubs/nchrp/docs/NCHRP21-06_FR.pdf


45

ACRONYMS AND ABBREVIATIONS

ASTM American Society for Testing and Materials AWWA American Water Works Association CI cast iron CR corrosion rate DI ductile iron DIP ductile iron pipe ECDA external corrosion direct assessment EPA U.S. Environmental Protection Agency LPR linear polarization resistance max maximum min minimum mpy mils per year PCA pipeline condition assessment PE polyethylene Ppm parts per million WERF Water Environment Research Foundation WRF Water Research Foundation µmhos/cm micro mhos per centimeter µs/cm micro Siemen per centimeter Ω*cm Ohm centimeter


47

APPENDIX A

UTILITY QUESTIONNAIRE

AND

SUMMARY OF UTILITY QUESTIONNAIRE RESPONSES


48

EVALUATING THE CURRENT CONDITION AND FUTURE PERFORMANCE OF DUCTILE IRON PIPE (DIP)

Tell Us About Yourself

Name: Title / Position: Utility: Street Address:

City: State: Zip:

Phone General: Direct: Cell: E-Mail: Web:

Tell Us About Your Utility

Piping Transmission miles - Diameter to inches Length Distribution miles - Diameter to inches Customer Residential Base Commercial

Daily Delivery mgd

Tell Us About Your System

System Bare Cast Iron Pipe miles Details Coated Cast Iron Pipe miles

Coating Materials Bare Ductile Iron Pipe miles Coated Ductile Iron Pipe miles

Coating Materials Other Pipe Materials

Coated Steel miles PCCP miles Other miles Other miles

Percent of System that is Bare Ductile Iron Percent

When did bare DIP use start? How many failures on bare DIP per year? Is this an increasing number or trend? Yes Do you have detailed leak records? Yes What information is obtained?

Age , Diameter , Location on pipe , Site location , Pit depth , Surface condition , Photos

Can they be reviewed? Yes


49

How does failure manifest itself? How is a pipe failure defined? Check all that apply:

External pitting , Internal pitting , Longitudinal split , Circumferential crack , Other Describe Other

Which type is most prevalent?

Do you have a corrosion control program? Yes Is this conducted by in-house group? Yes Is this conducted by consultant? Yes Do you use cathodic protection? Yes

Sacrificial Yes Impressed Current Yes Coating Yes

Tell Us About Your Environment

Soil Corrosivity: Very Corrosive , Moderately Corrosive , Mildly Corrosive , Non-corrosive

Type: Clay , Silt , Loam , Sand , Other Describe Other

pH: Acidic , Neutral , Alkaline

Are there Stray Current concerns? Yes

What is the total cost associated with a leak? Base Repair Cost $ Property Damage $

Other: $ $

Tell Us About Your Testing

Any planned excavations in the near future? Yes Why are excavations being conducted?

Can soil samples be collected for analysis? Yes

What information do you obtain at a repair site?

Do you have pit depth gages? Yes Do you have voltmeters & reference electrodes? Yes


50

Will you participate in data collection? Yes Are you interested in database participation? Yes Will you review/comment on field procedures? Yes

How do you see your in-kind participation dollars being allocated?


51

TABLE 1: Questionnaire Responses Regarding Piping System, Client Base and System Details

TABLE 2: Questionnaire Responses Regarding Leak Information and Future Testing

Piping System

ResponseTransmission Length (mi)

Distribution Length (mi)

Residential (k)

Commercial(k)

Daily Delivery

(kgd)

Bare Cast Iron (mi)

Coated Cast Iron

(mi)

Bare Ductile

Iron (mi)

Coated Ductile

Iron (mi)

Percent System

Bare DIP

Number Failures on

Bare DIP per Year

Sum 4111 26226 2137 186 1056 12889 76 9517 1760 150Max 1859 4055 396 34 180 2000 76 2091 878 83% 107Min 8 1026 47 5 21 62 0 0 0 0% 0

Average 316 2017 16 14 75 921 6 732 160 38% 28

NOTE: Based on 14 utilities

Customer Base System Details

Leak Information Future Testing

Response

Increaseing Number

or Failure Trend

Detailed Leak

Records

External Pitting

CC Program

Stray Current

Planned Excavations

Collect Soil Samples

Pit Depth Gage

DVM & RE

Data Collection

Yes 6 14 11 7 11 9 12 3 5 12No 6 0 1 7 3 5 2 11 9 2

NOTE: Based on 14 utilities


DISCUSSION

Tables 1 and 2, above, present a final summary of the Utility Questionnaire Responses received. They were used to generate the following graphs. The summaries are based on the 14 received utility responses.

Figure 1: System Composition based on 14 responses

Figure 1 presents the breakdown of the water system between transmission and distribution piping. It is clear that these operators are dealing with significantly more distribution piping than transmission piping as is usually the case. (It is believed that one value is skewing the ranges of both transmission and distribution piping based on a response that only had a transmission piping value.)

The same information was reviewed again and compiled by utility (listed by number so their identity is not revealed) so the amount of piping in each transmission and distribution system could be determined. This information is provided in Table 3, below, and then presented graphically in Figure 2. The table clearly shows some problems with the responses. For example, Utility 8 has indicated all transmission piping and has not broken out distribution piping while Utility 12 has done the opposite and reported all piping as distribution piping (which is actually easier to believe).

52©2015 Water Research Foundation. ALL RIGHTS RESERVED.

53

TABLE 3: Questionnaire Responses By Utility Regarding System Composition

Utility Transmission Length (mi)

Distribution Length (mi)

1 117 11172 214 10263 8 12174 80 11945 221 28026 210 40007 423 27138 18599 193 129110 363 405511 12 1486 13 145 118314 163 2429

15 115 1713

Figure 2: System Composition by Utility based on 14 responses


54

Figure 3 provides information regarding the utilities’ client base. It shows the number of customers both residential and commercial clients as well as the amount of water delivered daily.

Figure 3: System Delivery based on 14 responses

Piping system details are presented in Figure 4. It is interesting to note the difference in the amount of coated pipe between cast iron and ductile iron pipe. (All operators understood that “bare” pipe comes from the fabricator with a mill-applied asphalt coating. “Coated” pipe was any post fabricator applied system like tape wrap or polyethylene encasement.)


55

Figure 4: System Details based on 14 responses

It was felt that the information presented in Figure 4 would be more appropriate if viewed by each utility. In this manner each utility’s piping system could be examined. Table 4, below, provides the responses from each utility. The table clearly shows some problems with the responses. For example, Utility 8 reported only 62 miles of bare cast iron (and no other piping) but earlier reported 1859 miles of combined transmission and distribution piping. Utility 12 reported 1246 miles of bare cast iron (less than the total distribution miles reported above of 1486 miles) and no ductile iron pipe, bare or coated, even though they reported they started using DIP in 1965.

Figure 5, below, presents the information from Table 4 as a stacked bar graph so that the system details for each utility can be seen.


56

TABLE 4: Questionnaire Responses By Utility Regarding System Details

Utility Bare Cast

Iron (mi)

Coated Cast Iron (mi)

Bare Ductile

Iron (mi)

Coated Ductile

Iron (mi)

1 375 0 593 1 2 800 0 310 0 3 180 0 1019 0 4 418 0 376 214 5 999 t0 1660 0 6 2000 0 320 620 7 1829 0 213 878 8 62 0 09 511 0 870 0 10 1750 2091 11 12 1246 0 13 1098 76 0 47 14 772 0 1424 0

15 849 0 641 0

Figure 5: System Details by Utility based on 14 responses


57

Figure 6 presents the response range for percent of bare ductile iron pipe in the utilities’ systems as well as the number of failures on bare DIP each year.

Figure 6: Percent Bare and Number of Failures based on 14 responses

Figure 7 shows the responses regarding leaks and failures. All respondents indicated that they have detailed leak record but only half indicated that they had a corrosion control program. Yet eleven utilities indicated that they are concerned with stray currents.

As shown in Figure 8, excavations are planned by nine utilities and twelve agreed to collect soil samples and also gather data. But a few important components will be missing from the field gathered data since only three utilities have pit depth gages and only five have digital voltmeters for taking pipe-to-soil potential measurements and pH readings. We need to consider how we can turn their inability to collect all data but their willingness to participate to our advantage.


58

Figure 7: Leak Information based on 14 responses

Figure 8: Future Testing based on 14 responses


59

APPENDIX B

AQUAMATE™ PORTABLE CORRATER® INSTRUMENT


62


63

APPENDIX C

LPR PROBE AND MODIFICATIONS


64

1. Original LPR Probe

The original LPR probe was based on work done under an old NCHRP project. Theold drawings, shown below, were modified to use ductile iron rings rather than carbonsteel and only measure LPR.

Figure 2: Original NCHRP LPR Probe Exploded View

Figure 3: Original NCHRP LPR Probe Tip


65

2. Prototype LPR Probe

The resulting drawing was used to fabricate the prototype probe shown below.

Figure 4: Original LPR probe fabricated for WaterRF.

Figure 5: Overall view of manufactured LPR Probe


66

Figure 5: Close up of probe tip with two ductile iron rings and three insulating spacers

Figure 6: Close up of probe handle and wiring.

3. Final LPR Probe

This probe was modified to make it easier to use in the field. The probe was reducedto 1-inch diameter. The larger diameter required too much working of the probe to getit into the soil and then contact with the surrounding soil is questionable.

The wire connection was brought out through the side of the handle rather than ontop. There was a concern that leaning onto the top of the probe handle may damagethe wires.


67

A consensus was reached after discussions with industry representatives and panel members that a probe with ductile iron rings containing an annealing oxide would be more representative to actual in service DI water mains. We coordinated with an iron pipe foundry and had the newly machined DI rings sent out for heat treating to create the desired annealing oxide.

The final resultant probe used during our laboratory testing phase is shown below (2 were fabricated).

Figure 7: Overall view of final manufactured probe.

Figure 8: Close up of probe tip with annealed ductile iron rings.


68


69

APPENDIX D

LABORATORY TESTING


70

Soil samples were received in Corrpro’s Houston Soils Lab and stored until LPR probe testing could be conducted at the same time as laboratory testing (Figure 1). Multiple readings were taken in the same bucket changing AquaMate meters each time. Readings were also taken over increasing time periods to verify that the probe stabilizes rather quickly. Lastly, probes were placed side by side and measurements were performed. Table 1, below, presents the corrosion rate data measured over time.

Figure 1: Received soil samples stored in Houston.

Figure 2: Side-by-side LPR probe testing.


71

Table 1: Corrosion Rate Bucket Testing Data Over Time Bucket Testing

Sample Number Time

Probe 1 Corrosion

Rate

Probe 2 Corrosion

Rate 10 2 min 0.30 0.27 10 10 min 0.28 0.27 11 2 min 0.03 0.11 11 10 min 0.03 0.13 12 2 min 0.22 0.32 12 10 min 0.23 0.33 19 2 min 0.39 0.61 19 10 min 0.44 0.61 20 2 min 0.04 0.18 20 10 min 0.05 0.19 35 2 min 0.78 0.73 35 10 min 1.25 0.79

The corrosion rate data measured in side-by-side testing was compared to the laboratory measured corrosion rate in a two-electrode soil box with thick DI plates. This data is presented in Table 2, below. It is easy to see in the data that the two-electrode box yielded higher corrosion rates than the LPR corrosion rate probes. There is currently not enough data to determine the exact ratio between the two approaches. Several outliers skew the graph and possible multiplier. More data in the future may reduce the impact of the outliers and show a tighter correlation.

Imbalance was also analyzed between the two LPR probes. Table 3 presents the average CR and average imbalance for the probes. It also presents the same information for the two-electrode soil box with thick DI plates. While the readings between the bucket probe data and the lab testing data are not exactly the same, they do demonstrate the same trend; higher corrosion rate correlates to a higher imbalance or pitting potential.

Resistivity data was calculated from the AquaMate solution conductivity data obtained for each probe. The data is presented in Table 4. The same data for each soil sample was obtained from the two-electrode soil box using the AquaMate Meter and Nilsson Soil Resistance Meter. The AquaMate solution conductivity data was converted to resistivity.

Soil sample chloride content and pH data determined through laboratory testing are presented in Table 5.

All data is summarized and presented in Table 6.


72

Table 2: Corrosion Rate Data


Sample Number

Probe 1 Corrosion

Rate

Probe 2 Corrosion

Rate

Probe Average

CR DI Thick Plates CR

1 0.76 1.77 1.27 0.872 0.37 0.39 0.38 0.423 1.02 1.29 1.16 0.654 0.63 0.33 0.48 0.865 1.37 1.42 1.40 1.896 2.30 3.14 2.72 3.587 0.52 0.80 0.66 2.298 0.16 0.11 0.14 0.439 0.88 0.44 0.66 1.8410 0.28 0.27 0.28 0.9811 0.08 0.04 0.06 0.3112 0.23 0.33 0.28 1.0913 0.34 0.10 0.22 0.1414 0.21 0.19 0.20 0.1215 0.62 0.56 0.59 0.1516 1.38 0.99 1.19 0.7617 1.83 1.92 1.88 4.9518 2.42 1.90 2.16 2.6919 1.14 1.58 1.36 3.1020 0.05 0.19 0.12 0.4721 0.25 0.11 0.18 0.1422 1.58 2.43 2.01 9.9123 2.06 3.12 2.59 8.8924 1.60 1.22 1.41 0.7325 1.87 2.13 2.00 3.2826 1.49 2.20 1.85 2.6327 1.73 1.98 1.86 5.6828 3.09 2.49 2.79 5.5929 1.75 2.25 2.00 2.4430 2.31 2.46 2.39 12.331 0.32 0.26 0.29 0.0233 0.19 0.08 0.14 0.2734 0.38 0.26 0.32 0.0635 2.24 1.86 2.05 3.0336 0.22 0.21 0.22 0.3137 0.15 0.07 0.11 0.19


73

Table 3: Corrosion Rate and Imbalance Data


Sample Number

Probe Average

CR (mpy)

Probe Average

Imbalance

DI Thick Plates

CR (mpy)

DI Thick Plates

Imbalance 1 1.27 6.51 0.87 2.852 0.38 1.69 0.42 0.773 1.16 5.32 0.65 1.294 0.48 2.61 0.86 6.595 1.40 7.28 1.89 0.026 2.72 12.95 3.58 0.497 0.66 3.28 2.29 0.398 0.14 0.28 0.43 0.199 0.66 4.26 1.84 2.1310 0.28 1.78 0.98 1.0211 0.06 0.51 0.31 0.7712 0.28 1.84 1.09 1.3513 0.22 1.09 0.14 0.0314 0.20 0.33 0.12 0.2315 0.59 5.24 0.15 0.8216 1.19 8.59 0.76 0.1617 1.88 8.10 4.95 6.3918 2.16 11.10 2.69 4.8219 1.36 8.37 3.10 1.4020 0.12 1.05 0.47 1.0821 0.18 0.83 0.14 1.5222 2.01 11.17 9.91 10.4023 2.59 12.00 8.89 6.3124 1.41 4.23 0.73 1.7825 2.00 7.88 3.28 5.9426 1.85 8.39 2.63 1.9127 1.86 10.02 5.68 2.8028 2.79 11.30 5.59 8.7229 2.00 13.60 2.44 4.3930 2.39 11.24 12.3 11.9031 0.29 0.64 0.02 0.2533 0.14 0.15 0.27 0.7234 0.32 0.61 0.06 0.1535 2.05 26.55 3.03 9.1036 0.22 0.27 0.31 1.4337 0.11 0.22 0.19 0.29


74

Table 4: Corrosion Rate and Resistivity Data

Sample Number

Probe 1 Corrosion

Rate(mpy)


Probe 2 Corrosion

Rate(mpy)


Probe Average


DI Thick Plates CR

(mpy)

DI Thick Plates


DI Thick Plates

Resistivity Nilsson

(ohm-cm)

1 0.76 1,969 1.77 2,212 2,090 0.87 3,040 3,2002 0.37 11,236 0.39 18,519 14,877 0.42 10,753 11,0003 1.02 4,630 1.29 5,319 4,974 0.65 6,410 6,0004 0.63 5,618 0.33 7,143 6,380 0.86 3,704 3,6005 1.37 1,905 1.42 3,040 2,472 1.89 3,610 3,4006 2.30 1,183 3.14 613 898 3.58 1,377 1,2007 0.52 11,628 0.80 12,987 12,307 2.29 2,451 2,3008 0.16 33,333 0.11 76,923 55,128 0.43 10,309 10,0009 0.88 6,667 0.44 10,000 8,333 1.84 2,660 2,50010 0.28 4,292 0.27 6,494 5,393 0.98 9,524 8,80011 0.08 34,483 0.04 125,000 79,741 0.31 18,182 16,00012 0.23 11,628 0.33 7,692 9,660 1.09 6,849 5,60013 0.34 24,390 0.10 100,000 62,195 0.14 90,909 84,00014 0.21 200,000 0.19 200,000 200,000 0.12 45,455 52,00015 0.62 2,364 0.56 2,545 2,454 0.15 5,435 6,80016 1.38 4,405 0.99 5,780 5,093 0.76 5,128 9,60017 1.83 3,257 1.92 2,890 3,074 4.95 3,690 3,60018 2.42 437 1.90 887 662 2.69 883 48019 1.14 2,237 1.58 2,710 2,474 3.10 3,448 2,90020 0.05 142,857 0.19 32,258 87,558 0.47 27,027 13,00021 0.25 23,256 0.11 111,111 67,183 0.14 7,246 7,60022 1.58 1,238 2.43 1,056 1,147 9.91 1,805 1,30023 2.06 3,030 3.12 2,985 3,008 8.89 3,378 2,70024 1.60 5,882 1.22 3,497 4,689 0.73 4,902 3,10025 1.87 4,405 2.13 3,333 3,869 3.28 3,185 2,40026 1.49 6,024 2.20 6,579 6,302 2.63 9,346 36,00027 1.73 3,311 1.98 4,049 3,680 5.68 5,525 6,00028 3.09 812 2.49 833 823 5.59 1,193 1,10029 1.75 1,548 2.25 1,645 1,596 2.44 2,304 1,70030 2.31 556 2.46 576 566 12.3 832 80031 0.32 32,258 0.26 45,455 38,856 0.02 200,000 18,00033 0.19 83,333 0.08 500,000 291,667 0.27 52,632 36,00034 0.38 50,000 0.26 40,000 45,000 0.06 90,909 72,00035 2.24 790 1.86 1,984 1,387 3.03 1,342 1,10036 0.22 142,857 0.21 333,333 238,095 0.31 43,478 72,00037 0.15 500,000 0.07 500,000 500,000 0.19 52,632 56,000



75

Table 5: Corrosion Rate, Chloride Content and pH Data

Bucket Testing

Lab Testing

Sample Number

Probe Average CR

(mpy)

DI Thick Plates CR

(mpy) pHChlorides

(ppm)

1 1.27 0.87 8.3 42 0.38 0.42 7.0 23 1.16 0.65 6.9 44 0.48 0.86 7.9 605 1.40 1.89 7.7 426 2.72 3.58 7.8 1607 0.66 2.29 7.9 4208 0.14 0.43 7.7 209 0.66 1.84 7.3 84

10 0.28 0.98 7.3 28

11 0.06 0.31 7.2 2

12 0.28 1.09 7.9 33

13 0.22 0.14 6.9 1

14 0.20 0.12 7.7 4

15 0.59 0.15 6.3 4

16 1.19 0.76 6.3 7

17 1.88 4.95 6.3 6

18 2.16 2.69 7.4 420

19 1.36 3.10 7.0 15

20 0.12 0.47 4.7 14

21 0.18 0.14 8.1 10

22 2.01 9.91 7.1 4

23 2.59 8.89 7.7 96

24 1.41 0.73 5.0 160

25 2.00 3.28 7.7 33

26 1.85 2.63 4.6 10

27 1.86 5.68 4.6 8

28 2.79 5.59 5.1 290

29 2.00 2.44 7.4 84

30 2.39 12.3 7.6 250

31 0.29 0.02 7.1 3

33 0.14 0.27 6.6 12

34 0.32 0.06 7.7 1

35 2.05 3.03 4.4 6

36 0.22 0.31 5.3 1

37 0.11 0.19 5.0 1



76

TABLE 6: Summary of Data

Sample Number

Probe Average

CR (mpy)

Probe Average

Imbalance

Probe Average


DI Thick Plates CR

(mpy)

DI Thick Plates

Imbalance

DI Thick Plates


DI Thick Plates

Resistivity Nilsson

(ohm-cm)

Moisture Content

% pHChlorides

(ppm)Sulfides (ppm)

Standard Electrode

Conductivity (µmhos/cm)

Standard Calculated Resistivity (ohm-cm) Sample Type

Soil Classification*

DI Thick Plates CR

(mpy)

DI Thick Plates

Imbalance

DI Thick Plates


DI Thick Plates

Resistivity Nilsson

(ohm-cm)

1 1.27 0.5 2,090 0.87 2.85 3,040 3,200 28.0 8.3 4 0 610 1,600 Clay loam Corrosive 1.34 4.38 2,110 2,1002 0.38 1.00 14,877 0.42 0.77 10,753 11,000 16.0 7.0 2 0 170 5,900 Clay loam Mildly Corrosive 1.41 0.40 5,848 5,2003 1.16 1.50 4,974 0.65 1.29 6,410 6,000 28.0 6.9 4 0 450 2,200 Sandy loam & clay Moderately Corrosive 1.26 0.77 2,994 2,7004 0.48 2.00 6,380 0.86 6.59 3,704 3,600 10.0 7.9 60 0 1,000 1,000 Silty clay Corrosive 3.18 2.80 2,381 2,0005 1.40 2.50 2,472 1.89 0.02 3,610 3,400 25.0 7.7 42 0 390 2,600 Silty clay Moderately Corrosive 2.06 1.49 3,049 2,5006 2.72 3.00 898 3.58 0.49 1,377 1,200 17.0 7.8 160 0 1,500 670 Sandy clay Very Corrosive 3.65 1.33 1,350 1,2007 0.66 3.50 12,307 2.29 0.39 2,451 2,300 14.0 7.9 420 0 2,300 440 Silty loam & rocks Very Corrosive 2.64 1.41 1,048 1,0008 0.14 4.00 55,128 0.43 0.19 10,309 10,000 10.0 7.7 20 0 670 1,500 Sandy clay Corrosive 1.47 0.35 3,096 2,9009 0.66 4.50 8,333 1.84 2.13 2,660 2,500 19.0 7.3 84 0 1,000 1,000 Sandy clay Corrosive 3.06 3.02 1,976 1,700

10 0.28 5.00 5,393 0.98 1.02 9,524 8,800 23.0 7.3 28 0 400 2,500 Sandy clay loam & rocks Moderately Corrosive 1.44 0.11 5,263 5,200

11 0.06 5.50 79,741 0.31 0.77 18,182 16,000 17.0 7.2 2 0 120 8,300 Sandy clay loam Mildly Corrosive 1.04 0.53 8,696 7,600

12 0.28 6.00 9,660 1.09 1.35 6,849 5,600 22.0 7.9 33 0 530 1,900 Sandy clay loam & rocks Corrosive 1.95 0.93 2,825 2,500

13 0.22 6.50 62,195 0.14 0.03 90,909 84,000 3.8 6.9 1 0 120 8,300 Fine sand Mildly Corrosive 0.93 0.99 18,868 16,000

14 0.20 7.00 200,000 0.12 0.23 45,455 52,000 5.3 7.7 4 0 380 2,600 Silty loam & rocks Moderately Corrosive 1.46 0.05 4,608 4,400

15 0.59 7.50 2,454 0.15 0.82 5,435 6,800 16.0 6.3 4 0 600 1,700 Sandy clay loam Corrosive 1.43 5.13 1,783 1,600

16 1.19 8.00 5,093 0.76 0.16 5,128 9,600 19.0 6.3 7 0 310 3,200 Sandy clay Moderately Corrosive 3.45 4.29 3,497 3,000

17 1.88 8.50 3,074 4.95 6.39 3,690 3,600 34.0 6.3 6 0 660 1,500 Silty clay Corrosive 6.95 6.26 2,551 2,700

18 2.16 9.00 662 2.69 4.82 883 480 18.0 7.4 420 0 3,000 330 Sandy clay Very Corrosive 4.13 3.73 744 380

19 1.36 9.50 2,474 3.10 1.40 3,448 2,900 36.0 7.0 15 0 300 3,300 Rocks & clay loam Moderately Corrosive 2.71 2.59 3,367 2,900

20 0.12 10.00 87,558 0.47 1.08 27,027 13,000 15.0 4.7 14 0 150 6,700 Rocks & clay loam Mildly Corrosive 1.10 0.51 10,309 8,000

21 0.18 10.50 67,183 0.14 1.52 7,246 7,600 11.0 8.1 10 0 220 4,500 Rocks & clay loam Moderately Corrosive 1.61 1.74 4,082 3,200

22 2.01 11.00 1,147 9.91 10.40 1,805 1,300 28.0 7.1 4 0 4,100 240 Sandy clay loam Very Corrosive 14.6 10.9 1,049 1,000

23 2.59 11.50 3,008 8.89 6.31 3,378 2,700 20.0 7.7 96 0 650 1,500 Sandy loam Corrosive 8.75 7.94 2,404 2,000

24 1.41 12.00 4,689 0.73 1.78 4,902 3,100 25.0 5.0 160 0 820 1,200 Silty clay loam Corrosive 2.78 2.08 1,639 1,400

25 2.00 12.50 3,869 3.28 5.94 3,185 2,400 45.0 7.7 33 0 480 2,100 Silty loam & rocks Moderately Corrosive 3.26 5.96 3,040 2,400

26 1.85 13.00 6,302 2.63 1.91 9,346 36,000 43.0 4.6 10 0 110 9,100 Silty clay loam & rocks Mildly Corrosive 2.73 1.23 9,091 36,000

27 1.86 13.50 3,680 5.68 2.80 5,525 6,000 19.0 4.6 8 0 410 2,400 Clay loam Moderately Corrosive 5.88 2.83 5,495 5,600

28 2.79 14.00 823 5.59 8.72 1,193 1,100 39.0 5.1 290 0 1,400 710 Clay Corrosive 5.99 10.3 1,192 920

29 2.00 14.50 1,596 2.44 4.39 2,304 1,700 19.0 7.4 84 0.4 700 1,400 Clay Corrosive 2.85 3.76 2,299 1,700

30 2.39 15.00 566 12.3 11.90 832 800 21.0 7.6 250 0 1,400 710 Clay Corrosive 11.7 9.46 835 800

31 0.29 15.50 38,856 0.02 0.25 200,000 18,000 10.0 7.1 3 0 150 6,700 Sandy clay loam Mildly Corrosive 1.15 1.45 9,524 7,600

33 0.14 16.50 291,667 0.27 0.72 52,632 36,000 16.0 6.6 12 0 160 6,300 Sand & clay Mildly Corrosive 3.12 2.04 12,658 14,000

34 0.32 17.00 45,000 0.06 0.15 90,909 72,000 8.0 7.7 1 0 100 10,000 Sand & rocks Mildly Corrosive 0.83 0.71 23,256 16,000

35 2.05 17.50 1,387 3.03 9.10 1,342 1,100 19.0 4.4 6 0 1,700 590 Clay Very Corrosive 3.16 5.55 1,495 960

36 0.22 18.00 238,095 0.31 1.43 43,478 72,000 11.0 5.3 1 0 49 20,000 Sandy clay loam Non-Corrosive 2.55 2.10 17,241 15,000

37 0.11 18.50 500,000 0.19 0.29 52,632 56,000 10.0 5.0 1 0 33 30,000 Sandy clay loam Non-Corrosive 0.54 0.37 26,316 19,000

*FHWA-NH1-00-44, Corrosion Degradation of Soil Reinforcements for Mechanically Stabilized Earth Walls and Reinforced Soil Slopes

Saturated Sample

Bucket Testing Lab Testing Standard Laboratory Testing


77

APPENDIX E

FIELD DATA COLLECTION FORMS

AND

LPR PROBE PROCEDURES


78

Field Data Collection Protocol


79

BARE DUCTILE IRON PIPE (DIP) INSPECTION REPORT FIELD DATA COLLECTION PROTOCOL

1. Assign an Incident Number. This Number is any unique code assigned by on-sitepersonnel so the Inspection Report, photographs and soil samples can be tracked.

a. Example: SWU 120530A – Springfield Water Utility on May 30, 2012,Location A

2. Anytime “Other” is selected, please provide additional information.3. Provide as much information regarding the pipe’s history, its service location, local

and adjacent surroundings.4. Photograph general location of surroundings from grade level.

a. Identify photographs with the Incident Number and sequential numbering.i. Example: Photo SWU 120530A – 1, Photo SWU 120530A – 2, etc.

5. Photograph surface of the pipe in excavation site prior to cleaning.6. Prepare the pipe surface for examination by clearing and cleaning soil and corrosion

product from the pipe surface. The surface may be cleaned with water orcompressed air assisted with a wire brush and geologist’s hammer. Do not use anyopen blades or knives.

7. Describe what you see after cleaning the pipe surface.a. Where was the pipe damaged?b. What did the corrosion product look like?c. Describe the surface of the pipe.d. Where was the leak located, if applicable?e. How did the pipe fail, if applicable?f. What potentially caused the failure?g. What repair technique was used on the pipe?h. What did the soil look like?i. Was the soil wet or dry (if pipe leaked or surface cleaned with water, make

determination based on observations during preliminary excavation oradjacent areas in ditch)?

8. Photograph general view of pipe surface after cleaning.9. Photograph close up of corrosion sites and pitting.

a. Use the figure on Page 3 to hand sketch areas photographed. The top of thepipe is viewed as the 12 o’clock position. Looking in the direction of flow,positions around the pipe correspond with clockwise movement. Distancealong the pipe is approximate based on total length excavated.

10. Where equipment is available, pH Reading is obtained by placing a copper/coppersulfate electrode and an antimony electrode side by side in the soil connected to adigital voltmeter and measuring the potential difference or voltage (DC).

11. Where equipment is available, Pipe-to-Soil Potentials are measured by making apositive connection to the pipe (use a test lead clip or awl) and negative connectionto a copper/copper sulfate electrode and measuring the potential difference orvoltage (DC). The electrode may be placed in the soil adjacent to the pipe at eachend of the ditch where the excavation ends. The recorded potential may be a stablevalue or fluctuating value in stray current areas. Record the appropriate value(s).


80

12. Collect two soil samples in two separate, one-gallon zip lock bags. Use permanentwaterproof marker to note Incident Number on the outside of the bag followed by –1 or – 2.

a. Example: Soil SWU 120530A – 1, and Soil SWU 120530A – 2.13. Where equipment is available, measure pit depths using a pipe pit depth gage. Place

the measuring instrument on the pipe surface parallel to direction of flow. Use thefigure on Page 3 to hand sketch areas measured. Identify and record the 3-5 deepestpits in each corroded area in the space provided on Page 4.

14. Send completed Field Data Collection Forms via USPS or E-mail to:Corrpro Companies, Inc. Attention: Emer Flounders 1800 Saint Georges Road Dresher, PA 19025 (215) [email protected]

15. Send soil samples to Corrpro’s Houston Soils Lab:Corrpro Companies, Inc. Attention: Nancy Jacob 7000B Hollister Houston, Texas 77040 (713) 460-6042


81

Field Data Collection Sheet


82

BARE DUCTILE IRON PIPE (DIP) INSPECTION REPORT

Incident #: Utility: Date: By:

Reason for Excavation: Leak Repair Explain Other:

Pipe Size: inch Class: Installed/Age: / Pipe Manufacturer: Service: Transmission Location / Street: City: Zip: GPS: Latitude Longitude Cover: Pavement Depth: feet of cover Adjacent Traffic: Light (Local raod or driveway)

Explain Other: Describe Surroundings:

Prepare surface for examination by clearing and cleaning soil and corrosion product from the pipe surface. Use a wire brush and geologist’s hammer. (Safety Note: Do not use any open blades or knives.)

Damaged Area: Pipe Barrel Corrosion product: Rusty

Explain Other: Surface Description: Generally corrosion free

Explain Other: Leak Location: o’clock position on pipe Failure Mode: Circumfirential Break

Explain Other: Cause: Third Party

Explain Other: Repair Technique: Leak Clamp

Explain Other: Soil Type: Clay

Explain Other: Adjacent Soil: Wet


83

Incident #:

SOIL TESTING AND COLLECTION pH Readings (Antimony Electrode)

pH Potential: mV Converted to pH:

Pipe-to-Soil Potentials (Copper/Copper Sulfate Electrode) Steady P/S Potential: mV Stray Current Area P/S Potential: Max mV, Min mV, Avg mV

Obtain two soil samples from the bottom of the excavation site: Collect soil in one gallon zip lock bag and seal the bag (If plastic storage

containers are used, seal the container with plastic tape) Use permanent waterproof marker to note Incident #, above, on the outside

of the bag followed by a -1 or -2 Send soil samples to Corrpro’s Houston Soils Lab:

Corrpro Companies, Inc. Attention: Nancy Jacob 7000B Hollister Houston, Texas 77040 (713)460-6042

PHOTOGRAPHS AND PIT DEPTH MEASUREMENT Obtain photographs of the following:

1. General location shots of surroundings from grade level2. Surface of the pipe in excavation site prior to cleaning3. General view of pipe surface after cleaning4. Close up of corrosion sites and pitting

Obtain pit depth measurements using a Thorpe Pipe Pit Gage or a Dial Depth Pit Gage, as shown below.

Figure 8: Thorpe Pipe Pit Gage Figure 2: Dial Depth Pit Gage

Tools must be placed on pipe surface parallel to direction of flow


84

Incident #:

FIELD SKETCH

Length = feet

12:00

3:00

6:00

9:00

12:00


85

Incident #:

PIT DEPTH MEASUREMENTS


86

Sample Field Data Collection Form


FIELD

87

BARE DUCTILE IRON PIPE (DIP) INSPECTION REPORT

Incident #: SWU 120530A Utility: Springfield Water Utility Date: May 30, 2012 By: Tom Jones

Reason for Excavation: Leak Repair, Break Repair, Maintenance, Other Explain Other:

Pipe Size: 12 inch Class: 350 Installed/Age: 1990 / 22 Pipe Manufacturer: American Service: Transmission, Distribution Location / Street: 1250 Washington Ave City: Springfield State: MO Zip: 09026 GPS: Latitude Longitude Cover: Pavement, Road, Gravel, Open Lot, Earth Depth: 4 feet of cover Adjacent Traffic: Light (Local road or driveway), Medium (Thru street),

Heavy (Main street w/trucks & buses), N/A (Behind curb or in grass), Other

Explain Other: Describe Surroundings: Suburban area. Light adjacent traffic.

Prepare surface for examination by clearing and cleaning soil and corrosion product from the pipe surface. Use a wire brush and geologist’s hammer. (Safety Note: Do not use any open blades or knives.)

Damaged Area: Pipe Barrel, Bell, Spigot, Joint, Valve Corrosion product: Rusty, Black, Odor/Smells, Other, None

Explain Other: Surface Description: Generally corrosion free, Uniform corrosion, Localized

corrosion forming random pits, Tightly adherent corrosion product, Other Explain Other:

Leak Location: N/A o’clock position on pipe Failure Mode: Circumferential Break, Longitudinal Crack, Pitting, Gasket

Failure, Joint, Corroded Bolts, Corporation, Tap, Other Explain Other: N/A

Cause: Third Party, Corrosion, Settlement, Traffic Vibration, Hinging on Object, Wash Out, Other Explain Other: N/A

Repair Technique: Leak Clamp, Strap, Replace section, Other Explain Other: None

Soil Type: Clay, Silt, Loam, Sand, Rocky, Other Explain Other:

Adjacent Soil: Wet, Moist, Dry


FIELD

88

Incident #: SWU 120530A

SOIL TESTING AND COLLECTION pH Readings (Antimony Electrode)

pH Potential: mV Converted to pH: 6.2

Pipe-to-Soil Potentials (Copper/Copper Sulfate Electrode) Steady P/S Potential: -625 mV Stray Current Area P/S Potential: Max mV, Min mV, Avg mV

Obtain two soil samples from the bottom of the excavation site: Collect soil in one gallon zip lock bag and seal the bag (If plastic storage

containers are used, seal the container with plastic tape) Use permanent waterproof marker to note Incident #, above, on the outside of the

bag followed by a -1 or -2 Send soil samples to Corrpro’s Houston Soils Lab:

Corrpro Companies, Inc. Attention: Nancy Jacob 7000B Hollister Houston, Texas 77040 (713)460-6042

PHOTOGRAPHS AND PIT DEPTH MEASUREMENT Obtain photographs of the following:

1. General location shots of surroundings from grade level2. Surface of the pipe in excavation site prior to cleaning3. General view of pipe surface after cleaning4. Close up of corrosion sites and pitting

Obtain pit depth measurements using a Thorpe Pipe Pit Gage or a Dial Depth Pit Gage, as shown below.

Figure 9: Thorpe Pipe Pit Gage Figure 2: Dial Depth Pit Gage

Tools must be placed on pipe surface parallel to direction of flow


89


FIELD SKETCH

Length = 5 feet

12:00

3:00

6:00

9:00

12:00

TOP

BOTTOM

TOP

DIRECTION OF FLOW

Location 1

Photo SWU 120530A - 2

Location 2

Photo SWU 120530A - 3


90


PIT DEPTH MEASUREMENTS

LOCATION 1 LOCATION 2 0.149 0.139 0.122 0.126 0.119 0.122


91


Figure SWU 120530A – 1: General View Figure SWU 120530A – 2: Location 1, Pitting on surface


92


Figure SWU 120530A – 3: Location 2, Pitting on surface near end


LPR Probe Procedures

93


LPR Probe Procedures

1. LPR Probe

1.1. Verify discontinuity between the 2 ring electrodes and discontinuity betweeneach ring electrode and the probe shaft.

Figure 1: Discontinuity check – ring to ring

Figure 2: Discontinuity check – ring to shaft

1.2. Clean tip with moist rag but do not use abrasives.

94


2. LPR Meter

2.1. Check battery, perform test probe readings and check settings on meter.2.1.1. Settings for meter:

2.1.1.1. Multiplier = 1.32 (See below for calculations) 2.1.1.2. Cycle = 4 2.1.1.3. Element = standard 2.1.1.4. Probe = 2 electrode 2.1.1.5. Advanced = no

Figure 3: Meter check with test probe adaptor

3. Field Procedures

a. Insert rod into soil (1-2 feet minimum) ensuring good contact of soil around tip. It may be necessary to create a pilot hole in soil (Figure 5). Pilot hole should be made with bar of smaller diameter than the LPR rod to ensure good contact of soil around the tip. Probe may also be inserted into the excavation trench wall running parallel to the pipe.

95


Figure 4: LPR field probe equipment

Figure 5: LPR probe tip and small diameter pilot hole rod.

96


Figure 6: LPR probe inserted into soil and meter connected.

3.2. Hook up meter to LPR probe. Once the probe is set do not touch or move the probe during readings.

3.3. Take readings at 30 seconds, 2 minutes and 10 minutes. 3.4. Record all data.

4. Laboratory Soil Bucket Procedure

4.1. Same as Field Procedure, above but in a 5-gallon bucket. Emphasis on ensuring good circumferential soil contact around the LPR probe tip.

NOTE (Multiplier Calculations) New Probe SA = pi D L

L = 0.188 in x 2.54 cm/in = 0.47752 cm

D = 1 in x 2.54 cm/in = 2.54 cm

SA = 3.14159 2.54 0.47752 = 3.81044 cm^2

New Probe Multiplier= 5 / 3.81044 = 1.31219

Use 1.32

Quad Box SA = pi r^2 0.50

r = 2.544 cm

= 3.14159 6.472 0.5 = 10.1661 cm^2

Quad Box Multiplier= 5 / 10.1661 = 0.49183

Use 0.5

97©2015 Water Research Foundation. ALL RIGHTS RESERVED.

development and testing of a linear polarization

Documents