Improved Test Methods and Practices for Characterizing Steel Corrosion Potential of Earthen Materials (2020)

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

7 Chapter 1. Background 1.1. Introduction Electrochemical properties of earthen materials such as electrical resistivity, pH, salt concentrations and organic contents are commonly used to characterize their corrosion potential. AASHTO test standards, adopted in the early 1990s, are among the most common practices in the United States to determine the electrochemical properties of earthen materials. However, the AASHTO test methods do not apply to all earthen materials that encompass a broad range of physical and electrochemical characteristics, nor do they distinguish issues inherent to particular types of infrastructure construction. For example, AASHTO T-288 (2016) is used to determine the minimum resistivity (ρmin) of earthen materials at a saturated or slurry state. However, a slurry state does not represent a condition that occurs during the effective service lives of earth retaining structures. The minimum resistivity obtained from such a test is not representative of the resistivity of earthen materials experienced at any time during the service lives of metal elements placed within them. Also, AASHTO T-288 (2016) specifies how test specimens are prepared by separating the sample into fractions according to particle size and only including the fraction passing the No. 10 sieve in the test specimen. Resistivity measurements for earthen materials are affected by soil texture, thus including only the finer portion of the sample within the test specimen renders resistivity measurements that are different from what would be measured if all of the particle sizes inherent to the sample were included. The AASHTO tests may not be appropriate for determining the corrosivity of coarser types of earthen materials and do not consider practical limits on moisture contents that may be experienced in the field. Results from the current AASHTO practices cannot be interpreted beyond establishing a common reference point to compare the corrosivity of different soils under laboratory conditions. This research evaluates alternative test methods that may be more appropriate for particular applications (e.g., MSE Walls), and will consider a wider range of fill types incorporating larger particle sizes. 1.2. Research Objectives The main goal of this research is to develop a test protocol for characterizing the in-service steel corrosion potential of earthen materials. The objectives needed to achieve this goal include: (1) identifying, sampling, and characterizing representative earthen materials; (2) determining the effects of different electrochemical measurement techniques and different specimen preparation procedures (e.g., aggregate size) on the measured electrochemical properties of compacted soil or leachates extracted from the solid samples; (3) establishing links between laboratory and field measurements for proper interpretation of laboratory test results; and (4) developing a test protocol and corresponding characterization of corrosion potential that more accurately reflects the corrosivity of earthen materials compared to the conventional methods. In pursuit of these objectives, we addressed a number of questions and technical challenges, including: a. Quantifying sampling and testing errors associated with measurements of electrochemical properties considering: a. diversity in the site and environmental conditions, earthen materials used in construction, and construction practices, and b. variations in test procedures to measure electrochemical properties.

8 b. Evaluating how the compositions of earthen materials and water chemistry affect measurements of electrochemical properties considering: a. porosity, mineralogy, and texture (tortuosity) of earthen materials, and b. the relevance of laboratory tests, test parameters and sample preparation techniques to the actual field conditions/applications. c. Relating laboratory measurements of electrochemical properties to performance observed from field measurements, considering: a. how well sampling strategies and the end points for laboratory measurements (e.g., the final moisture content) apply to the specific field conditions; b. material characterizations that are consistent with the available field performance data; c. the reliability of laboratory and field tests; d. service life design and asset management practices for transportation infrastructure; and e. barriers to implementation. 1.3. Review of Current Practices for the Characterization of Corrosion Potential of Earthen Materials This section summarizes the current test methods used to evaluate the corrosivity of earthen materials and presents the suggested interpretation of results. Laboratory tests for measurements of a) electrical resistivity b) chloride content c) sulfate content, and d) pH were included in the investigation. Sample preparations and limitations associated with each test method are also presented in this section. 1.3.1 Factors Affecting the Corrosion Potential of Geomaterials Several electrochemical parameters influence the corrosivity of an earthen material, including electrical resistivity, degree of saturation, pH, dissolved salts (ions), and redox potential (Elias 1990). These properties can also be affected by contamination from constituents not typically components of soil including less common minerals from mining, and contamination from natural petroleum or manmade fertilizers. The effects from contaminants were not directly included in this study. Most salts are active participants in the corrosion reaction with the exception of carbonate, which forms an adherent protective scale on the surface of most metals and inhibits corrosion rate. According to the literature, chloride, sulfate, and sulfide are the major components promoting corrosion in steel reinforcements embedded in earthen materials/concretes (Ahmad 2003; Romanoff 1957). Given its relationship to salt content, Romanoff (1957) and King (1977) established resistivity as the most significant indicator of corrosion potential in earthen materials. Sagues et al. (2009) identified the following factors affecting corrosion of metals in soils and water: 1) Key factors are temperature, oxygen concentration, resistivity, pH, carbonate scaling tendency, acids, alkalis, salts, soil particle size distribution, porosity, water content, and microbial activity. 2) The maximum corrosion occurs at a critical moisture content in a soil mass., Above the critical moisture content corrosion rate is controlled by the conductivity of the soil/water mixture (i.e., activation controlled or the rate by which electrons from the metal are transferred to oxygen

9 molecules through the electrolyte). Below the critical moisture content corrosion rate is diffusion controlled (i.e., controlled by the rate of oxygen diffusion towards the metal surface). 3) Generally, higher annual rainfall and higher temperatures produce groundwater that is highly corrosive (as temperature increases, the ion mobility and corrosivity increases). 4) Carbonate scaling can contribute to reduced corrosion rates, so the presence of carbonates should be considered for determining corrosivity of soils and water. 5) Other properties of water such as dissolved oxygen content may need to be considered for precise determination of corrosion potential. 6) Besides chloride, sulfate can also break down the protective passive film (or carbonate scaling) and cause pitting corrosion (William 2009). Hence, observations of the performances of metals in contact with earthen materials with higher sulfate contents, but low chloride contents, are needed to improve characterizations of corrosion potential. 7) The observed elevated corrosion rates at sites not initially characterized as aggressive due to the soil/water properties were mainly attributed to microbial induced corrosion (MIC). MIC is expected to be more significant in marine environments, warmer climates, and for metal in contact with soils that have a high organics content. MIC can be mitigated via use of engineered, free draining fill that is free of organics, and placement of metals in soils above the water table where there is an abundance of oxygen. 1.3.2 Current Laboratory Test Methods Most transportation agencies evaluate electrochemical properties of earthen materials using current AASHTO laboratory test standards, which were adopted in the early 1990s. Specified test methods to measure the electrochemical properties of geomaterials are AASHTO T-288 (2016) (resistivity measurement), AASHTO T-289 (2018) (pH measurement), AASHTO T-290 (2016) (soluble sulfate content measurement), and AASHTO T-291 (2013) (soluble chloride content measurement). AASHTO describes the electrochemical requirements for fill material to be suitable for MSE wall construction as shown in Table 1-1. Table 1-1 AASHTO requirements for the fill materials in MSE walls. Parameter Acceptable range AASHTO standard Minimum resistivity > 3000 Ω-cm T-288 (2016) pH 5 – 10 T-289 (2018) Sulfate content < 200 ppm T-290 (2016) Chloride content < 100 ppm T-291 (2013) Alternatives to AASHTO tests to measure soil resistivity include ASTM G187 (2018), Tex-129-E (1999), Tex-129-M, ASTM WK 24621, and SCT 143 (2008). Resistivity test methods are of two general types that include (1) measurements of voltage drop in response to an applied current passing through specimens that are compacted in a soil box (galvanostatic test), or (2) conductivity measurements on aqueous solutions extracted from solid samples (leachates). Other differences between the tests are in terms of sample treatments that may include sieving, air drying, heating, methods of mixing, time of settling/curing, and filtering. Test methods ASTM WK 24621 and Tex-129-M are new test methods (under development during the course of this research, 2016-2019) considered for implementation by ASTM and TxDOT.

10 Tex-129-E (1999) was the TxDOT standard that will be superseded by Tex-129-M. Tex-129-M is an improvement compared to Tex-129-E as it applies to testing coarse graded samples including gravel. For Tex-129-E larger particles were crushed to render specimens with all particles passing a #8 sieve. However, the geochemical behavior and electrochemical properties of crushed particles are not representative of the larger sized particles from which they were derived, and the electrochemical and geochemical activity that occurs on the surfaces of the larger particles (Bronson et al. 2013). The test procedure described in Tex-129-M allows all particle sizes up to a maximum of 1 ¾ inches to be include in the test specimen, and relative to Tex-129-E larger sized boxes are employed to accommodate testing coarse samples. ASTM WK 24621 also applies to coarse graded samples. In general, tests for pH and salt content are performed on extracts obtained after diluting a small solid sample with deionized (DI) water. Specific details of specimen preparation vary amongst the different test procedures such as the size of the solid sample, fraction of earthen material included in the test specimen (e.g., portion finer than No. 10 sieve), dilution ratio, soaking period, method and time of mixing, and filtration of solids. These factors can significantly affect the obtained electrochemical results. Alternatives to AASHTO tests for measurement of pH include ASTM D 4972 (2019), SCT 143 (2008), Tex-128-E (1999), Tex-620-M, a test procedure developed at CorrTest and described as part of NCHRP Project 21-06 (2009), and a new test method for determining pH of lightweight aggregates which was being considered by ASTM Committee D18 during the course of this research. The latter two test methods and Tex-620-M are applicable for measuring the pH of relatively coarse-grained materials, while the other tests are more applicable to finer materials. Alternatives to AASHTO tests to measure soluble salt contents include Tex-620-J (2005) and Tex- 620-M. In addition, ASTM D4327 (2017) provides a more robust technique which uses ion exchange chromatography (IC) to determine the soluble salt content. This technique can be applied to the samples that are prepared in accordance with AASHTO T-290 (2016) and AASHTO T-291 (2013). In addition, the sulfate and chloride contents can be determined from the same specimen. We reviewed state DOT standard specifications to identify their practices for measuring electrochemical properties of earthen materials and detailed results are presented in Appendix B. Twenty-two states use AASHTO test methods, three states referenced multiple test methods, one state publishes modifications to the AASHTO methods, 12 states do not use the AASHTO methods, 15 states use different electrochemical requirements, and one state references FHWA guidance instead of AASHTO specifications for their practice. 1.3.2.1 Comparison of Different Resistivity Test Methods Tests for measurement of soil resistivity include those performed on extracts, or on compacted specimens at moisture contents that include as-received and saturated. Resistivity measurements may be made in-situ or in the laboratory. Laboratory measurements have the advantage that the moisture content in the soil box is controllable. In order to obtain a comparable resistivity that is independent of seasonal and other variations in soil-moisture content, resistivity should be determined under the most adverse condition (e.g., at saturation). The resistivity measured in the

11 water-saturated soil box does not necessarily represent the actual site conditions, but provides a baseline for comparing the corrosivity of different earthen materials, and is considered as the minimum resistivity in this study. AASHTO T-288 (2016), ASTM G-187 (2018), Tex-129-E (1999), Tex-129-M and ASTM WK 24621 are test methods for measurements of resistivity on compacted specimens. Tests performed on compacted specimens are useful to investigate the influence of moisture content, level of compaction, and particle size distribution (i.e., tortuosity of the current flow path) on specimen resistivity. Figure 1-1 shows the typical process for resistivity measurements of earthen materials using a two- electrode soil box used in AASHTO T-288 (2016). A sample size that includes about 1500 g of air-dried materials finer than No. 10 sieve is required for testing. • The soil sample is placed in an acrylic plastic soil box (Figure 1-1 (a)) in layers and compacted using finger pressure (Figure 1-1 (b)). The soil box has inner dimensions of 150 mm × 100 mm × 45 mm (length × width × height). • In order to provide a proper electrical contact between the resistivity meter and the soil, two stainless steel plates with dimensions of 150 mm × 45 mm are affixed to the side walls of the soil box (distance between stainless steel electrodes = 100 mm). • A measured amount of distilled or DI water, with a resistivity greater than 20,000 Ω-cm, is gradually added to the soil sample. • The resistivity meter is then connected to the stainless-steel electrodes, as shown in Figure 1-1 (e), and an alternating current (AC) consisting of a square wave with a frequency of 97 Hz is passed through the soil sample. • The electrical resistance is measured from the corresponding voltage drop between the two electrodes. The resistivity of the soil sample is computed by multiplying the resistance by the soil box factor, which is a function of the geometry of the box. The process is repeated by remixing the soil sample with increasing amounts of distilled or DI water to produce resistivity measurements at various moisture contents (up to saturation state, shown in Figure 1-1 (d)). A plot of resistivity versus moisture content renders the minimum resistivity and corresponding moisture content (Figure 1-2).

12 Figure 1-1 Resistivity measurement using a two-electrode soil box. AASHTO T-288 (2016), ASTM G187 (2018), ASTM WK 24621, Tex-129-E (1999), and Tex- 129-M differ in terms of sample treatments (whether or not the sample is dried before distilled water is added in increments), the particle size distribution of the specimen, the manner in which the soil sample is mixed with water, and the moisture conditions during the test. In Table 1-2 we summarize the differences between different test methods in terms of sample treatment and specimen preparations. In the present research, the results from different test methods are compared to those from the relevant AASHTO standards, which serve as the nominal values. In case of soil resistivity, we make these comparisons with respect to data from the AASHTO T-288 (2016) test. Figure 1-2 Resistivity versus moisture contents (adapted from McCarter, W.J. (1984)). soil sample resistivity meter water saturation compaction weight measurement soil box (a) (b) (c) (d) (e) 0 500 1000 1500 2000 2500 0 5 10 15 20 25 30 35 R es is tiv ity ( Ω -m ) Moisture content (%) ρmin

13 Table 1-2 Comparison of different resistivity test methods in terms of sample treatment. Test methods Air/oven dry Particle size Mixing method Moisture condition AASHTO T-288 (2016) air/oven dried at 60⁰ C < 2 mm; crushing not allowed water added incrementally, mixed thoroughly, placed in box; 1st increment cures for 12 hrs. water added in increments until saturated or until reaching a minimum resistivity ASTM G- 187 (2018) No Debris and particles > ¼ inches removed unless tested as-received, water is added and mixed as soil is placed within the box; no curing as-received or saturated ASTM WK 24621 No, soaked for 24 hrs. prior to testing all sizes similar to ASTM G-187, but aggregates have been soaked as received/saturated then drained Tex-129-E (1999) oven-dried at 60⁰ C < 2.36 mm; crushing allowed water added incrementally, mixed thoroughly, placed in box; no curing water added in increments until saturated Tex-129-M air/oven dried at 60⁰ C all sizes water added incrementally, mixed thoroughly, placed in box; no curing water added in increments until saturated Test methods performed on aqueous extracts for measurement of resistivity/conductivity include Tex-620-M, the USGS Field Leach Test (Hageman 2007) and SCT 143 (2008). Leachate tests commonly include (a) preparing measured amounts of material for testing, (b) adding a measured volume of DI water to the sample, (c) agitating the mixture (simultaneous heating in some of the test methods), and (d) measuring pH, temperature, and conductivity of the aqueous solution (see Figure 1-3). The samples are syphoned via syringes and filtered before conducting the analytical tests using IC to determine the sulfate and chloride contents. The USGS Field Leach Test (USGS FLT) applies to poorly graded sands, gravels, and aggregate mixtures. This test measures the conductivity of leachate and has the advantage that the same sample may be used for measurements of resistivity, pH, chloride, and sulfate contents. However, this test may not render meaningful results for well graded materials, where the tortuosity of the path between the particles affects the current flow and the obtained resistivity measurement. Different test methods such as SCT 143 (2008), Tex-620-J and Tex-620-M differ with respect to sample preparations that may include sample size, dilution ratio (the weight ratio of DI water added to the soil sample), the manner in which samples are mixed with water, and whether or not the extract is filtered before the test. Table 1-3 summarizes the differences between these test methods in terms of sample preparation.

14 Figure 1-3 Procedure to perform a typical leach test. (a) prepare measured amounts of material, (b) add DI, (c) agitate, (d) make measurements Table 1-3 Differences in test methods performed on extracts (leachates). Test methods Sample size (g) (W:S) 1 Mixing method Settling time (hr.) Filtration SCT 143 2000 1:1 Mixed then stand for 30 minutes, agitate for 3 minutes at 0, 2- and 4-hour intervals 20 Yes Tex-620-J 30 (separated on the #4 sieve and then pulverized to pass a #40 sieve) 10:1 Heat sample to 150°F and digest on hot plate for approximately 16 hours stirring periodically none Yes Tex-620-M 100 (dried) 10:1 shake vigorously for 30 minutes 1 No; tip of electrode placed 5 cm deep into the mixture 1 W:S = water to solids ratio (by weight). a b c d

15 1.3.3 Limitations of Current Test Methods AASHTO T-288 (2016), T-289 (2018), T-290 (2016) and T-291 (2013) are performed on specimens that are separated from the sample on the No. 10 sieve. In particular, for coarse fills with little to no material passing the No. 10 sieve, a sufficient amount of fines for testing might be obtained from sieving a large quantity of the material. This process is impractical and may lead to inaccurate results, especially for gravel fills that have very little fine materials. Unless some breakage is anticipated during placement and compaction, crushing of larger aggregates to obtain the finer fraction is neither appropriate, nor allowed by the AASHTO test standards. This is because most of the soluble ions are concentrated on the surfaces of the particles (diffusions of ions through the particles are negligible). Testing the finer portion (passing a No. 10 sieve, i.e., finer than 2.00 mm) of the material, presumes that the finer fractions are significant sources of soluble salts. This is not necessarily the case when using coarse fills that have very little, or no, material finer than the No. 10 sieve. In that case an alternative method of test should be considered. 1.4. Knowledge Gaps and Study Purpose Gaps in knowledge that need to be addressed before we can recommend alternative test methods for measurements of electrochemical properties are summarized with the following questions: • How fine does the material need to be before testing the fraction passing the No. 10 sieve is appropriate? • How coarse should the material be before testing an aqueous extract for resistivity is appropriate? • How do results obtained from different test methods compare? • What is the precision and bias for individual test methods? • How well does the proposed characterization of corrosion potential correlate with performance/observed corrosion rates?