Electrochemical Test Methods to Evaluate the Corrosion Potential of Earthen Materials (2021)

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.

51 4.1 Introduction During Phase III of NCHRP Project 21-11 the research team cooperated with selected trans- portation agencies that implemented the recommended protocol as a shadow specification. The data include characterization of different sample sources (e.g., maximum particle size and gradation) along with measurements of the geochemical and electrochemical properties of the samples, including resistivity, pH, and chloride and sulfate content (i.e., the total salt content). The Wenner four-probe technique, according to ASTM G57 and Wenner (1915), was used in the field for measurement of electrical resistivity, and representative material samples were collected from the site or from the source in order to perform electrochemical tests in the labo- ratory with the Texas modified and AASHTO test procedures. The source is where materials are stockpiled before they are used for construction; samples from the site are taken after these materials have been transported, placed, spread, and compacted for the project. Sampling in this manner provides a comparison between materials that are sampled and tested before con- struction versus after construction. Differences between the two samples are possible because of variations within the source or alterations to the materials that occur as they are transported, placed, spread, and compacted. This chapter summarizes the field and laboratory test results obtained from four active con- struction projects in cooperation with two owners (the Texas and New York State DOTs), four different geotechnical testing laboratories (the University of Texas at El Paso, McMahon & Mann, the New York State DOT, and the Texas DOT), and four general contractors. The main focus of Phase III was to determine whether the main technical goal of the research project— better characterization of corrosion potential through using the suggested protocol as compared with the traditional methods—had been met. The practicality and implementation of the sug- gested protocol through the interaction with laboratories engaged in electrochemical testing and with suppliers/owners in different states was also evaluated. This chapter presents a brief description of the data obtained in pursuit of Phase III. This is followed by the key results obtained from different test methods in the form of resistivity as mea- sured by the Wenner four-probe test (used in the field) and laboratory measurements, including resistivity, pH, and chloride and sulfate content. A detailed discussion of the trends observed within the data sets is presented and the field and laboratory measurements are compared. 4.2 Description of Data Set Two sites in Texas and two sites in New York were included in the field verification study: San Antonio and El Paso, Texas, and Buffalo and Schroon, New York. Materials were sampled from the sources and at the sites for laboratory testing. Split samples were sent to the University C H A P T E R 4 Field Measurements

52 Electrochemical Test Methods to Evaluate the Corrosion Potential of Earthen Materials of Texas at El Paso and to state DOT geotechnical laboratories to determine gradation, resis- tivity, pH, and salt and sulfate content. Table 4-1 summarizes the laboratory tests performed on the samples obtained from the site and the source by the owner and/or the research group at the University of Texas at El Paso. The site in San Antonio involved construction of MSE walls at the intersection of Interstate 10 East and Ackerman Road (Figure 4-1a), where the MSE wall fill was a coarse, open-graded gravel produced locally by Hanson Aggregates (Lehigh Hanson Company 2019). Limestone was mined at the quarry to produce the aggregates for MSE wall fill. Similar to the San Antonio site, the site in El Paso involved construction of MSE walls at the intersection of Montana Avenue and Lee Trevino Drive (Figure 4-1b). The MSE fill at this site was a medium-grained, well-graded sand. Two quarries from the New York State inventory were selected for inclusion in the evalua- tion: one from the eastern part of the state (Peckham Quarry, New York State DOT Region 1) and one from the western part (Enterprise Stone & Lime, New York State DOT Region 5). The sources from the eastern and western parts of the state were granite and limestone quarries, respectively. The two sources produced well-graded gravels that included significant sand components. These materials meet the requirements for pavement subbase in New York State. The material sourced from New York State DOT Region 5 was used in a bridge reconstruc- tion project in Tonawanda, New York, near Buffalo along Interstate 290 west where the high- way crosses the Niagara Frontier Transportation Authority property (Figure 4-1c). The fill was placed as subbase beneath the shoulder backing at the west end of the bridge. The shoulder backing was compacted to a thickness of approximately 12 in. and was approximately 5 ft wide. Although this material was used in an application where corrosivity was not an issue, the New York State DOT sampled and tested the material similar to the way specified for MSE wall fill. Material from New York State DOT Region 1 was also used in a bridge reconstruction project along Interstate 87 north, near Exit 28 in Schroon (Figure 4-1d). Although this material could potentially be a source of MSE wall fill, for this project, material was placed as subbase beneath the shoulder along the approach sections. The material was compacted to a final lift thickness of approximately 15 in. and had been in place for less than a week during the research team’s site visit. Figure 4-2 presents gradation curves for the different materials sampled from the sites cited in this chapter. Table 4-2 summarizes the salient details of each sample in terms of aggregate size, grade number, and USCS classification to describe the characteristics of the sample domain used in the field and laboratory tests. Given the details from Table 4-2—grading number (GN) Performer of Test Sample Origin Site Source Owner (state DOTs) —a Measurements of ρ, [Cl–] and [SO4] content, and pH using Texas modified and AASHTO procedures Research group (UTEP) Texas modified and AASHTO proceduresb Texas modified and AASHTO procedures a No tests were performed by the owners on the site samples. bField resistivity tests were performed with the Wenner four-probe technique, according to ASTM G57. Table 4-1. Laboratory tests performed on samples from the site and source.

Field Measurements 53 Location of the shoulder 42°59’49”N 78°51’34.5”W 42.996944, - 78.859583 Location of MSE wall 31°48'03.1"N 106°18'52.8"W 31.80086428, - 106.31466715 Location of the MSE wall 29°26’18.7”N 98°22’48.2”W 29.438528, -98.380056 (a) (c) (b) (d) Location of the shoulder 43°52’16.3”N 73°45’13.9”W 43.871194, -73.753861 Figure 4-1. Locations and coordinates of construction sites: (a) San Antonio, (b) El Paso, (c) Buffalo, and (d) Schroon. (Photos from El Paso, Buffalo, and Schroon sites are older Google Earth images that do not depict construction activities at these locations.) 0 10 20 30 40 50 60 70 80 90 100 0.010.1110100 P er ce nt ag e fin er b y w ei gh t Particle size (mm) Sand FinesGravel San Antonio, TX Schroon, NY El Paso, TXBuffalo, NY Figure 4-2. Gradation curves from different sites.

54 Electrochemical Test Methods to Evaluate the Corrosion Potential of Earthen Materials and PP#10—and from the suggested protocol (appendix; see also Figure 3-14), samples from Buffalo and El Paso should be evaluated using the current AASHTO tests (particles passing the No. 10 sieve), and samples from San Antonio and Schroon should be evaluated using the Texas modified procedures, which include larger particles within the test specimens. 4.3 Results 4.3.1 Resistivity Measured by Soil Box Test Figure 4-3 shows the laboratory measurement of soil resistivity with a modified soil box for the samples retrieved from the sources and the sites included in Phase III. The equipment used for laboratory measurement of resistivity included a soil resistance meter (M. C. Miller Model 400D) and a potentiostat (Gamry Instruments Model 600). The potentiostat showed repeat- ability of about 10% to 15%, and the soil resistance meter showed repeatability of about 1%; therefore, the results obtained from the soil resistance meter are used in the following discussion. Site NMAS (in.) PP1/4 in. PP#10 PP#200 Cu Cc USCS Classification GN San Antonio, TX 2 0 0 0 1.48 0.97 GP 0.13 El Paso, TX 0.25 94 70 0 6.52 1.04 SW 4.55 Buffalo, NY 1 49 31 0 20.00 0.99 GW 3.07 Schroon, NY 2 17 10 0 11.01 2.51 GW 1.70 Note: NMAS = nominal maximum aggregate size; Cu = uniformity coefficient (D60/D10); Cc = coefficient of curvature [(D30)2/D60 × D10]. D10, D30, and D60 are the diameters corresponding to 10%, 30%, and 60% finer in the particle size distribution curves (gradation curves). Table 4-2. Summary of gradation from different sources. C1 C2 P2 P1 resistance meter 12 in. 8 in. depth = 3.5 in. stainless steel plates soil box (a) Gamry device M.C. Miller device DAQ soil box digital scale (b) Figure 4-3. Resistivity measurement with modified soil box: (a) schematic (C1 and C2 = current terminals; P1 and P2 = potential terminals) and (b) test setup (DAQ = data acquisition system).

Field Measurements 55 Figure 4-4 presents laboratory measurements of resistivity from samples retrieved from sources and sites in Buffalo, Schroon, San Antonio, and El Paso. This figure includes the results from testing samples according to AASHTO T 288 and Tex-129-M and data obtained by owners and contractors as well as by the University of Texas at El Paso (through the research team). The data presented in Figure 4-4 show relatively good agreement between the resistivity measure- ments from samples retrieved from the sources (tested by the University of Texas at El Paso and the owners) and those obtained from the sites (tested by the University of Texas at El Paso). An exception to this observation is the site sample from Schroon tested according to Tex-129-M (resistivity = 9,300 Ω · cm). This result appears to be low compared with the other test results obtained from the source and tested according to Tex-129-M and is not higher than the results obtained by following AASHTO T 288. This is unusual, as measurements of resistivity by Tex-129-M are generally higher than those obtained by AASHTO T 288 with the same sample, as depicted by the bias shown in Figures 3-4 and 3-5 and discussed in Section 3.3.1.2. 4.3.2 Resistivity Measured by Field Tests Using Wenner Technique In situ resistivity measurements were performed at each site using the Wenner four-probe technique, as described by ASTM G57 and Wenner (1915). In addition, in situ measurements of moisture content and density were made with a nuclear density gauge (Troxler Model 3440; see Figure 4-5), and samples were collected from the site for further laboratory investiga- tions. The following equipment was used for resistivity measurements: GEO Earth Ground Tester (Fluke Model 1625-2), soil resistance meters (Nilsson Model 400 and M. C. Miller Model 400D), and a potentiostat (Gamry Instruments Model 600). The resistivity values pre- sented in this research are mainly from those obtained with the M. C. Miller Model 400D soil resistance meter, as it provided the most stable resistivity readings. Figure 4-6 shows the typical test setup for performing the Wenner four-probe measure- ment. A penetration depth of 6 in. was used for all measurements, but varied spacing (6 in., Note: The owner of two of the El Paso and San Antonio sites (the Texas DOT) did not provide results of resistivity testing in accordance with Tex-129-M. The Texas DOT provided data from testing in accordance with AASHTO T 288 for the San Antonio site but not for the El Paso site. 2,400 2,200 2,100 8,600 12,000 10,700 20,600 17,300 16,700 4,000 3,300 4,400 6,400 6,300 23,000 36,000 9,300 45,300 61,300 4,700 4,300 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 O w ne r/ S ou rc e U T E P /S ou rc e U T E P /S ite O w ne r/ S ou rc e U T E P /S ou rc e U T E P /S ite O w ne r/ S ou rc e U T E P /S ou rc e U T E P /S ite O w ne r/ S ou rc e U T E P /S ou rc e U T E P /S ite Buffalo, NY Schroon, NY San Antonio, TX El Paso, TX R es is tiv ity ( -c m ) AASHTO T 288 Tex-129-M Figure 4-4. Comparison of resistivity results obtained from modified soil box tests.

56 Electrochemical Test Methods to Evaluate the Corrosion Potential of Earthen Materials 12 in., and 24 in.) was used between the probes. The spacing determines the depth covered by the measure ments. For resistivity measurements in MSE walls, test locations perpendicular and parallel to the reinforcement strips were used. For resistivity measurements along the shoulder (or shoulder back) of roads, measurement lines perpendicular and parallel to the pavement were used. Figure 4-7 is a schematic that depicts the locations, zones, and directions of resistivity measurements at each site. At locations where the soil appeared to be very dry (e.g., in the MSE wall in San Antonio), water was added in small amounts (less than 200 ml) to the soil next to the probes to improve the electrical conductivity between the probes and the soil. In this case, samples for measuring nuclear density gauge (Troxler Model 3440) Figure 4-5. In situ measurement of moisture content at Schroon. Soil C1 C2 P2 P1 Resistance meter a a a b Side view C1 and C2: current terminals P1 and P2: potential terminals (a) Wenner probes Gamry device M.C. Miller device DAQ (b) Figure 4-6. Wenner four-probe technique: (a) schematic and (b) test setup in the field.

Field Measurements 57 the moisture content and electrochemical properties of the soil were collected farther away from the wet region of the soil. These samples were sealed and stored in double Ziploc plastic bags to prevent evaporation and contamination. According to Wenner (1915), the resistance reading (R) can be converted to the resistivity (r) at different probe penetration depths (b) and different probe spacing (a) by using Equation 4-1. 4 1 2 4 2 4 4 (4-1) 2 2 2 2 aR a a b a a b r = π + + − + When the penetration depth of the probes is small in comparison to the spacing between them (b < 0.05 a), Equation 4-1 can be simplified as Equation 4-2, which is the formula cited by the ASTM G57 standard. Since the penetration depth in this research (6 in.) was clearly greater than 5% of the spacing between the probes, Wenner’s (1915) original formula (Equa- tion 4-1) was used to determine the numerical resistivity values in the field. 2 (4-2)aRr = π Figure 4-8 shows resistivity measurements as a function of depth for different sites at direc- tions parallel or perpendicular to the reinforcements (or pavements). These data include measurements from lines/locations spaced approximately 15 ft apart. Each line/location was (a) L1 L10 L10L1 L1 to L10 (every 15 feet) shoulder 2. 5 fe et – 3. 5 fe et perpendicular to the pavement parallel to the pavement 2 fe et 6 inches edge of pavement (reference point) N a (Typ.) (c) L1 L6 L1L2 L1 to L6 (every 15 feet) MSE wall fill 15 fe et perpendicular to the reinforcements parallel to the reinforcements 5 feet 2 feet edge of concrete panel (reference point) steel reinforcement 15 feet a (Typ.) N (b) Zone 1 Zone 2 a (Typ.)b a (Typ.) c probes perpendicular to the CL of the road probes parallel with the CL of the road Concrete slab Pavement N (d) L1 L5L5L1 L1 to L5 (every 10 feet) MSE wall fill perpendicular to the reinforcements parallel to the reinforcements 7 feet 1 foot edge of concrete panel (reference point) steel reinforcement a (Typ.) N Note: CL = centerline. Figure 4-7. Locations, zones, and directions for performing soil resistivity measurements: (a) San Antonio, (b) El Paso, (c) Buffalo, and (d) Schroon.

(a) 0 0.5 1 1.5 2 2.5 0 5,000,000 10,000,000 15,000,000 20,000,000 25,000,000 30,000,000 35,000,000 D ep th ( ft) Resistivity (Ω·cm) Location 1 Location 2 Location 3 Location 4 Location 5 Location 6 (b) 0 0.5 1 1.5 2 2.5 0 5,000,000 10,000,000 15,000,000 20,000,000 25,000,000 30,000,000 35,000,000 D ep th ( ft) Resistivity (Ω·cm) Location 1 Location 2 0 0.5 1 1.5 2 2.5 0 20,000 40,000 60,000 80,000 100,000 120,000 D ep th ( ft) Resistivity (Ω·cm) Location No.1 Location No.2 Location No.3 Location No.4 Location No.5 (c) 0 0.5 1 1.5 2 2.5 0 20,000 40,000 60,000 80,000 100,000 120,000 D ep th ( ft) Resistivity (Ω·cm) Location No.1 Location No.2 (d) (e) 0 0.5 1 1.5 2 2.5 0 20,000 40,000 60,000 80,000 100,000 120,000 D ep th ( ft) Resistivity (Ω·cm) Location 1 Location 2 Location 3 Location 4 Location 5 Location 6 Location 7 Location 8 Location 9 Location 10 (f) 0 0.5 1 1.5 2 2.5 0 20,000 40,000 60,000 80,000 100,000 120,000 D ep th ( ft) Resistivity (Ω·cm) Location 1 Location 2 Location 3 Location 4 Location 7 Location 8 Location 9 Location 10 (g) 0 0.5 1 1.5 2 2.5 0 20,000 40,000 60,000 80,000 100,000 120,000 D ep th ( ft) Resistivity (Ω·cm) Location 1 Location 2 Location 3 Location 4 Location 5 (h) 0 0.5 1 1.5 2 2.5 0 20,000 40,000 60,000 80,000 100,000 120,000 D ep th ( ft) Resistivity (Ω·cm) Location 2 Location 3 Location 4 Location 5 Figure 4-8. Results from in situ testing (determined by Equation 4-1): (a) San Antonio, parallel to reinforcements; (b) San Antonio, perpendicular to reinforcements; (c) El Paso, parallel to reinforcements; (d) El Paso, perpendicular to reinforcements; (e) Buffalo, parallel to pavement; (f) Buffalo, perpendicular to pavement; (g) Schroon, parallel to pavement; and (h) Schroon, perpendicular to pavement.

Field Measurements 59 a total of 6 ft long and included resistivity measurements at probe spacings of 0.5, 1.0, and 2.0 ft. At some locations, the 1- or 2-ft probe spacing could not be achieved because of space constraints. Figure 4-8 allows comparison of the data obtained from different sites and from fills that incorporate metal reinforcements and those that do not. The following observations were made on the basis of the data shown in Figure 4-8: ·  The resistivity measurements from the San Antonio site were very high—greater than 1,000,000 Ω · cm—and within the range of 5,000,000 to 35,000,000 Ω · cm. This was due to the extremely dry conditions at this site. ·  Data from the MSE wall site in San Antonio showed higher variation between measurements taken at different locations as compared with the data obtained from the other sites. This might be related to the extremely dry conditions at the site (moisture content less than 1%), as the operators had to add water next to the probes to the fill in small amounts to improve the electrical connection between the probes and the fill. Addition of water and the high evaporation rate at this site might be the main reasons for the relatively large variations in the resistivity measurements. ·  Measurements from the San Antonio and El Paso sites show higher resistivity values from lines oriented perpendicular to the MSE reinforcements as compared with the measurements from lines oriented parallel to the reinforcements. This can be explained by the fact that the steel reinforcement provides an easy path for current in the longitudinal direction, which causes a reduction in the resistivity reading when the alignment of the pins for resistivity measurements is oriented parallel (longitudinally) to the reinforcements. ·  The measurements from the New York sites that were performed within the shoulder area of the pavement show that measurements with lines perpendicular or parallel to the pavement are similar to each other (no steel reinforcements in the subbase materials). ·  In general, the resistivity measurements from the San Antonio site increase with respect to depth, but the data from the New York sites generally decrease with respect to depth. This observation may be related to the significantly higher effect of the reinforcements in reducing the resistivity at 0.5 ft depth compared with that at 2.0 ft depth (i.e., having one layer of reinforcement in a 0.5-ft thick material has a more significant charge-carrying effect as compared with one layer of reinforcement over a thickness of 2.0 ft). In addition, the measure ments at depths of 2.0 ft from the subbases studied at the New York sites reflect the presence of subgrade material beneath the 12–15 in. of subbase material that was placed for these projects (i.e., materials at the New York sites are not homogeneous over the 2-ft depth of measurements). ·  The resistivity measurements from the site in El Paso remained relatively constant with respect to depth, which is different from the increasing trend observed in the data from San Antonio. This could be attributed to the high moisture content of the fill in El Paso (average moisture content of 7.3%), which dominates over the effect of reinforcements (described above). In addition, the lower variation in the resistivity measurements from El Paso can be explained by the higher moisture content of the fill at this site. ·  Higher resistivity was observed in measurements at the site in Schroon as compared with those from the site in Buffalo. This was consistent with the laboratory measurements of resistivity. The in situ moisture contents at the Schroon, Buffalo, and El Paso sites were approximately 3.0%, 2.5%, and 7.3% by weight, respectively. Figure 4-5 shows one of the in situ moisture content measurements at Schroon performed by the New York State DOT (the owner) with a nuclear density gauge. The University of Texas at El Paso and the New York State DOT tested samples from the sources and sites in accordance with Tex-129-M, which includes the larger soil box and all of the particle sizes in the test specimen. The moisture-resistivity curves obtained

60 Electrochemical Test Methods to Evaluate the Corrosion Potential of Earthen Materials from these laboratory tests were extrapolated to moisture contents of 3.0%, 2.5%, and 7.3%, as shown in Figure 4-9, and these values were compared with the field/in situ measurements at depths of 6 in. As shown in Table 4-3, a good comparison was obtained. The average error was approximately 11%. 4.3.3 Other Electrochemical Properties Laboratory measurements of chloride and sulfate contents from samples retrieved from sources and sites in Buffalo, Schroon, San Antonio, and El Paso are presented in Figure 4-10 and Figure 4-11, respectively. These graphs include the results from testing samples according to AASHTO T 290, AASHTO T 291 and Tex-620-M. The data show good agreement between measurements from samples retrieved from the sources and those obtained from the sites. In most cases, the chloride and sulfate contents determined by Tex-620-M are lower than those obtained by AASHTO test procedures. An exception to this observation is seen with the source sample from Buffalo, for which the sulfate content obtained with Tex-620-M appears to be higher than that obtained with AASHTO T 290. However, the measurement of sulfate content obtained with AASHTO T 290 for the site sample is higher than the results obtained from testing the source. The measurement of sulfate content via AASHTO T 290 from the source sample appears to be anomalous and too low and was not considered in the comparisons of results. y = 239,758x-1.374 R² = 0.993 y = 112,488x-1.172 R² = 0.997 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000 0 5 10 15 20 R es is tiv ity ( ·c m ) Moisture content (%) at NYS DOT laboratory at MMCE ρ = 38,400 2.5% (in situ) Note: MMCE = McMahon and Mann Consulting Engineering and Geology. Figure 4-9. Extrapolation of moisture-resistivity curves obtained from laboratory tests to determine the resistivity at in situ moisture content for the site in Buffalo. Sample In Situ Range (Ω∙cm) Laboratory (Ω∙cm) El Paso, TX 8,000– 22,000 11,600 Buffalo, NY 30,000–54,000 38,400 Schroon, NY 60,000–115,000 95,000 Note: Comparisons between laboratory and in situ measurements for the San Antonio site are not included because of the unusually high resistivity measurements at this site (Figure 4-8, a and b). Table 4-3. In situ (Wenner four-probe) and laboratory (modified soil box) measurements.

Field Measurements 61 97 311 46 61 15 24 43 35 169 133 15 9 1 1 35 29 0 50 100 150 200 250 300 350 UTEP/Source UTEP/Site UTEP/Source UTEP/Site UTEP/Source UTEP/Site UTEP/Source UTEP/Site Buffalo, NY Schroon, NY San Antonio, TX El Paso, TX S ul fa te c on te nt ( m g/ kg ) AASHTO T 290 Tex-620-M Figure 4-11. Comparison of sulfate content results obtained from samples collected from the sites and sources. 16 11 5 26 8 4 54 45 8 5 1 1 1 1 47 33 0 10 20 30 40 50 60 UTEP/Source UTEP/Site UTEP/Source UTEP/Site UTEP/Source UTEP/Site UTEP/Source UTEP/Site Buffalo, NY Schroon, NY San Antonio, TX El Paso, TX )gk/g m( tnetnoc edirolh C AASHTO T 291 Tex-620-M Figure 4-10. Comparison of chloride content results obtained from samples collected from the sites and sources.

62 Electrochemical Test Methods to Evaluate the Corrosion Potential of Earthen Materials Table 4-4 shows the results of electrochemical tests performed on the samples retrieved from the sites and from the sources. The laboratory results of the resistivity tests are discussed in Section 4.3.1 (see Figure 4-4). Considering the AASHTO requirements for fills in MSE walls (Table 1-1), the material from Buffalo, is characterized as corrosive but other materials are noncorrosive. 4.4 Comments and Suggestions from Owners The research team exchanged information with the New York State DOT geotechnical and environmental engineering laboratories and noted their comments and suggestions for proper implementation of the proposed protocol. Because the Texas DOT participated in the develop- ment of the modified test procedures, it offered no additional comments. Following is a summary of the main comments and suggestion the team received from the New York State DOT: 1. When compacted specimens are being tested, do not add water directly to the compacted specimen within the soil box. Instead, add increments of moisture after the specimens are removed from the box and use a mixer similar to the one shown in Figure 4-12 to obtain a good distribution of moisture throughout the specimen. 2. The moisture increment corresponding to 100% saturation is easily obtained from weight measurements of the compacted soil specimen after each increment. The increment at which the wet weight of the material begins to decrease occurs near saturation. For AASHTO T 288 and Tex-129-M, compare the resistivity measurements obtained when the specimen reaches saturation (rsat) with the minimum resistivity (rmin) obtained by increasing the moisture content until the absolute minimum resistivity is reached. However, for moisture contents Sample Protocol Resistivity (Ω·cm) pH [Cl–] (mg/kg) [SO4] (mg/kg) Corrosive San Antonio, TX Modified tests 53,300 9.0 1 1 No El Paso, TX AASHTO 3,650 9.0 49 40 No Buffalo, NY AASHTO 2,233 7.7 14 311 Yes Schroon, NY AASHTO 10,433 8.3 15 53 No Note: The values in this table are the average of the site and the source. The shaded values do not meet current AASHTO requirements for MSE wall fill. Table 4-4. Characterization of corrosion potential. bucket containing the sample and added water electric motor Figure 4-12. Mixer used at Soils Engineering Laboratory of the New York State DOT.

Field Measurements 63 in excess of those needed to achieve 100% saturation, a slurry is tested rather than a com- pacted specimen. This comparison provides useful information to estimate the behavior of the material in worst-case scenarios and to determine the possible underestimation/ overestimation of the reported resistivity. 3. Preparation of coarse samples for resistivity tests in the 20-lb box and conducting the tests in accordance with Tex-129-M were feasible, and the amount of time and effort was reasonable. 4. A bottle roller is not standard equipment for a geotechnical lab, and other methods of mixing should be implemented for measuring salt contents and pH in accordance with Tex-620-M. 5. Mixing times for the tests on leachates for measurements of pH and salt contents should be kept to a maximum of 30 min. 6. Use a template to properly align pins for performing the Wenner four-probe test. 7. The Wenner four-probe test is relatively easy to implement and can be used on active con- struction projects. Resistivity meters built specifically for this test are recommended because of their greater ease of use, field ruggedness, and efficiencies. 4.5 Conclusions The following observations are drawn from the data collected in cooperation with owners/ contractors as part of Phase III: ·  The variations seen in results from laboratories that were testing similar samples were observed to be higher than the variations seen in resistivity measurements made with samples retrieved from the sources before construction and from the site during construction. This is because, in general, samples retrieved from the site or the source were similar. ·  The variations observed during the field studies included in Phase III were higher than the variations observed in testing replicates with the same operator/laboratory during Phase II. This is expected because of variabilities that exist between laboratory practices and errors due to variations between samples. The research team intentionally limited the variations between samples when the laboratory test program for Phase II was implemented, and this measure mitigated the sampling error. ·  Reinforcements affect in situ measurements of resistivity from MSE wall fills. ·  Reinforcements appear to have the least effect on measurements of resistivity if the lines for the Wenner four-probe test are oriented perpendicular to the reinforcements. ·  Good correspondence was observed between laboratory and field measurements of resis- tivity when the laboratory tests were conducted on compacted specimens with the same particle sizes, gradation, and moisture content as the in-place fill and considering the in-place moisture content of the fill. The experience and data collected from implementing the proposed protocol in active con- struction projects indicate that the modified test procedures and the test protocol for improved characterization of corrosion potential are easier to implement than the traditional methods. The owners/contractors were able to perform the modified test procedures and, with few exceptions, could acquire the equipment needed to perform these tests.