Advancements in Use of Geophysical Methods for Transportation Projects (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.

38 The survey results presented in Chapter 3 were used to select five agencies for further case examination. The agencies were primarily identified on the basis of responses that indicated the agency had frequent experience with geophysical methods or experience on noteworthy projects. The agencies selected were the Minnesota Department of Transportation (MnDOT), Ohio Department of Transportation (ODOT), Virginia Department of Transportation (VDOT), New Jersey Department of Transportation (NJDOT) and California Department of Transpor- tation (Caltrans). For each, the survey contact was interviewed and additional documentation was reviewed to document the history of geophysical use by the agency, common applications and noteworthy experiences, and helpful lessons on the application of geophysical methods for transportation projects. Minnesota Department of Transportation According to responses to the synthesis survey questions, the Minnesota Department of Transportation (MnDOT) is among the most prolific and experienced users of geophysical methods. MnDOT reported using geophysical methods at the highest rate—more than 10 times per year. The agency reported using 18 of the 36 geophysical methods listed in the survey, for nine of the 11 applications listed in the survey. MnDOT’s experiences with geophysical methods, including agency practices, commonly used methods, typical applications, and an example project, are discussed in the following sections. History of Geophysics Use and Agency Practice Geophysical methods became necessary for MnDOT after the agency experienced issues related to karst throughout Minnesota in the late 1990s. The agency found that the point-based approach of drilling alone to characterize the 3D subsurface structure was inadequate. As it con- tinued to deal with failures associated with karst features, such as subsidence and collapse within the MnDOT right-of-way, the agency looked to geophysical methods, particularly resistivity, to provide additional information about subsurface conditions. Over the past 20 years, MnDOT’s use of geophysics has expanded to other methods and applications. Geophysical methods are now used in a variety of applications that involve scoping, preliminary design, and assessment of postconstruction failures. MnDOT primarily uses in-house capabilities to perform geophysical measurements unless the project is designated design-build or construction manager/general contractor, in which case the geophysical measurements are performed by an external contractor. The agency has found that developing in-house capabilities for frequently used geophysical methods (such as resistivity, refraction, and surface wave methods) has been a cost-effective approach. C H A P T E R 4 Case Examples

Case Examples 39 A three-person group consisting of a geologist, geologic engineer, and geophysicist is responsible for collecting and analyzing geophysical data when the need arises. This group works closely with MnDOT’s drilling crews to provide supplementary information and fill in subsurface informa- tion between boreholes. In an interview, MnDOT personnel noted the importance of having borehole data to correlate and compare with the geophysical results; in some cases, though, as in the following case example, geophysics may be the only viable approach. MnDOT does not have specific guidelines for when to use geophysical methods. The selec- tion of geophysical methods is done on a case-by-case basis. Agency personnel weigh the quality of subsurface information available for the site, the expected quality of data for the specific application and field environment, the value of the information to be obtained from the measurement, and the cost of deploying equipment and processing the data. When select- ing the appropriate method for a site, MnDOT personnel also use a checklist of possible issues that could affect the results. For example, buried metal or power lines could affect electrical methods. Typical applications and environments in Minnesota where geophysical methods have proved effective are discussed in the next section. The availability of training resources has been important for MnDOT’s expansion of its in-house geophysics capabilities. Most training has been provided by equipment manufacturers. Training on how to process and interpret the geophysical data has been especially important. MnDOT personnel also report that short courses at conferences, such as surface wave courses at SAGEEP and other in-person training, have been beneficial. Commonly Used Methods and Applications MnDOT’s use of geophysics began with ER, which remains the most common and effective approach for many of the problems the agency encounters. Moist subsurface soil conditions make resistivity an especially effective method, while other approaches such as GPR are hampered by moist soil conditions that attenuate the signal. A common application of resistivity has been characterization of karst features such as cavities and voids, though it has also been used to determine general soil-rock stratigraphy, bedrock depth, and groundwater elevation, and to locate buried structures (Richter 2010). The most common application of ER in Minnesota has been to identify the locations and constrain the extent of organic soil deposits. MnDOT reports that ER provides a clear contrast between organic muck and the surrounding granular soils. Use of ER for this application has helped curb construction-phase costs for unplanned or excessive organic excavations. MnDOT also reports effective use of ER for forensic investigations, such as the investigation of the failure mechanism in an embankment constructed with Geofoam. In addition to ER, MnDOT performs several other methods in-house, including MASW, IP, SP, refraction, and downhole and crosshole measurements. MnDOT often runs IP in tandem with ER measurements and uses SP less frequently. For seismic methods, MnDOT reports that MASW is its go-to method for applications on the roadway, while refraction is typically used off the roadway for applications such as determining depth of bedrock. In the phone interview, MnDOT personnel noted how important it is for users of geophysical data to understand the methods’ capabilities and limitations. With better understanding of the methods, users can better determine if a measurement will provide the type of data they want. Geophysical methods have intrinsic limitations in terms of sensitivity and resolution, and are affected by environmental factors, such as cultural noise; geologic complexity; and other factors, such as the presence of power lines and buried metal. Unreasonable expectations of method capabilities may lead to application of the methods in unsuitable conditions. Knowledgeable practitioners and a good screening process can help avoid these problems.

40 Advancements in Use of Geophysical Methods for Transportation Projects MnDOT Example Project: Bridge over Miller Creek in Duluth, Minnesota An early application of ER by MnDOT took place in 2006 in Duluth, Minnesota, at the site of a new bridge, roadway, and retaining walls. The decision to use a shallow or a deep founda- tion for the new bridge at this site required information about the depth to bedrock. Although bedrock is typically around 20 ft deep in this area, nearby borings showed bedrock at 40 ft. The description that follows is taken from the Richter (2010) paper and the phone interview with MnDOT personnel. Access to the proposed bridge location for site investigation was difficult. Property owners denied access because of concerns about property damage and because Miller Creek, which runs through the area, is a designated trout stream. Because of these restrictions, MnDOT determined that traditional soil and rock borings and cone penetration test soundings could not be used. Instead, MnDOT used its newly acquired ER system to obtain subsurface information. Soil con- ditions within the bridge area were complex, with variably saturated silty, sandy, and gravelly organic soil and many boulders present (Richter 2010). Two ER surveys were performed, the first using 28 electrodes with 1-m spacing and the second performed orthogonally to the first using 56 electrodes with 1-m spacing. In this case, the resistivity results were interpreted without the benefit of boreholes for corroboration. The results generally showed a large contrast in resistivity at about 10 ft, which was attributable to the change between the more conductive near-surface soils and the underlying bedrock (Figure 28). Noisy data in the western portion of the survey produced artifacts that complicated the interpretation (Figure 28, bottom). Region C in Figure 28 (top) indicates a low-resistivity region in the interpreted bedrock, which was thought to be either a glacial pothole or a zone of highly weathered or fractured bedrock. The high-resistivity regions A and B in Figure 28 (top) were thought to be knobs of the gabbro bedrock. The survey results were communicated to personnel from the bridge office and the Duluth district. Before the survey, the default foundation option had been to use drilled piles because no information was available to support other options. The results from the resistivity survey, how- ever, provided support for using shallow foundations on the east side of the creek for the retaining walls and bridge abutment and possibly on the west side. Shallow foundations were designed for Figure 28. Resistivity results along north-south line (top) and east-west line (bottom) showing shallow bedrock starting at about 10 ft (courtesy of MnDOT).

Case Examples 41 the east end of the abutment, and deep foundation elements were designed for the west end with the option to switch to shallow foundations. When excavations were performed for the foundation, bedrock was encountered within 10 ft of the surface on both the east and the west side of the creek, which precluded the need for the deep foundation elements originally designed for the west side. The excavation also offered an opportunity to ground-truth the resistivity results. The exposed excavation showed two bedrock knobs (A and B) and a pocket of weathered and highly fractured bedrock (C), as shown in Figure 29. Primarily because of differences in material and equipment costs, MnDOT saw a modest cost savings of between $50,000 to $100,000 by constructing shallow foundations (Richter 2010). This project was the first time that MnDOT had based a bridge design solely on geophysical data. Because of its novel approach, the project was given an award for bridge construction by the General Contractors of America. More importantly, the project instilled in MnDOT personnel confidence in their use of geophysical methods. MnDOT: Lessons Learned MnDOT uses geophysical methods more than most state transportation agencies in the United States. The agency has found it cost-effective to perform most of its geophysical work in-house with on-the-job training supplied primarily by equipment manufacturers. Electrical resistivity has proved to be the most useful method for the conditions the agency faces, and MnDOT has had notable success using ER to define karst features and to delineate the extent of organic soil deposits for excavation. In the phone interview, MnDOT personnel emphasized the importance of know- ing the capabilities and limitations of the methods and of having a screening process to ensure that methods are used for the appropriate applications and environmental conditions. Person- nel also emphasized the need to corroborate the geophysical results with soil and rock borings whenever possible. The case example project presents a unique and successful case of an agency using ER results exclusively to design bridge foundations where soil and rock borings could not be performed. The ER results also showed strong agreement with ground truth from the excavation. Virginia Department of Transportation The Virginia Department of Transportation (VDOT) was also among the seven agencies that reported the highest frequency of geophysical method use (more than 10 times a year). VDOT also reported using 13 of the 36 geophysical methods listed in the survey, for 6 of the 11 applications listed in the survey and for one application that was not listed. Addition infor- mation about VDOT’s use of geophysical methods was gathered from a phone interview and review of relevant documents. VDOT’s experiences with geophysical methods, including agency practices, commonly used methods, typical applications, and details of a project, are presented in the following sections. History of Geophysics Use and Agency Practice Geophysics has been part of VDOT’s toolbox for about 30 years, but only since about 2010 has geophysics become a routine part of the subsurface investigation program. Unlike MnDOT, which performs many geophysical measurements in-house, VDOT’s current practice is to use geophysics through professional on-call contractors, who evaluate many of the agency’s promi- nent projects. With the exception of GPR, all geophysical methods at VDOT are performed by external contractors. When geotechnical services are procured, VDOT requires that geotechni- cal consultants either have the capability to perform geophysical measurements themselves or have subcontractors who can perform the work. Figure 29. Photo of excavated site associating features with two high- resistivity regions (A and B) and low- resistivity region (C) in resistivity results (courtesy of MnDOT).

42 Advancements in Use of Geophysical Methods for Transportation Projects Chapter 3 of the agency’s Manual of Instructions for the Materials Division (2019), Section 303.02, states that “geophysical exploration is an appropriate adjunct to a subsurface exploration program.” As stated in this manual and confirmed in the phone interview, VDOT requires that the methods score a 3 or 4 according to the U.S. Army Corps of Engineers manual EM 1110-1-1804 Geotechnical Investigations (2001) and are “appropriate for use on VDOT projects.” Appropriate use of geophysical methods is approved by the district materials engineer before implementation of the program. Commonly Used Methods and Applications VDOT’s use of geophysical methods has largely been driven by karst issues encountered in several areas of the state. The Staunton district in particular deals extensively with karst on many of its projects. Electrical methods, particularly electrical resistivity, have been the most commonly used in karst terrain because of their cost. In the phone interview, VDOT personnel reported generally good experiences with these methods but emphasized the need to perform confirmation borings to calibrate and corroborate the results. During the phone interview, VDOT personnel mentioned that electrical methods are commonly applied in situations where borings spaced 200 ft apart show very different subsurface profiles. In these cases, geophysical methods are often used to fill in between the boreholes and aid in selecting locations to perform addition borings. Agency personnel also mentioned that they have observed cases where geophysical methods produced inaccurate or ambiguous results; thus, the personnel cautioned against an overreliance on geophysical measurements in karst terrain. An example case is presented in the following section. VDOT personnel emphasized the benefits of using multiple geophysical methods at the same site, and also reported successful use of acoustic and optical televiewers on some projects. The agency reported some use of refraction measurements, and less frequent use of other seismic methods such as ReMi and MASW. The agency’s Manual of Instructions for the Materials Division (2019) emphasizes the use of geophysical methods to help characterize subsurface conditions in areas with intermediate geomaterials and suggests the following methods may be appropriate for this application: electrical resistivity, electromagnetic methods, refraction, reflection, GPR, microgravity, and surface wave methods. In both the manual and the phone interview, the importance of understanding the strengths and weaknesses of various techniques was emphasized. VDOT personnel also discussed in the phone interview the need for training resources on the types of geophysical methods available and their capabilities and limitations. VDOT Example Project: Inaccurate Prediction of Bedrock Depth on Bridge Project VDOT provided an example of a project where a complex resistivity image provided an inaccurate interpretation of bedrock depth. The project involved construction of a geosynthetic reinforced soil bridge, where the depth to bedrock at the site was of particular interest. Resistivity profiles were developed along three cross-sections that intersected borehole locations. The interpreted profile from one of the resistivity lines that passed through the location of borehole 14BH-002 and near borehole 14BH-001 is shown in Figure 30. Using the resistivity profile shown in Figure 30, the depth to bedrock at the location of 14BH-001 was interpreted to be about 23 ft (indicated by the dashed line). However, the boring 14BH-001 encountered bedrock at a shallower depth of 14.5 ft, where the resistivity interpretation indicated resistivity values consistent with wet clay soil. Although this borehole was offset 18 ft from the

Case Examples 43 Figure 30. Resistivity image with generalized interpretation and location of boreholes 14BH-001 and 14BH-002 (courtesy of VDOT). resistivity line, another resistivity line (not shown) passed though the 14BH-001 location and showed a similar predicted bedrock depth. The rock core from 14BH-001 is shown in Figure 31a. The error was more significant at boring 14BH-002, where the resistivity image indicated bedrock at about 17 ft and the boring showed bedrock at 35 ft. The rock core from 14BH-002 is shown in Figure 31b. A thin, 0.5-ft layer of limestone was encountered at 21 ft but was underlain by 14 ft of very stiff fat clay (CH). In this case, the CH material was interpreted as bedrock in the resistivity image. These results illustrate that the interpretation of resistivity is not unique and can provide inaccurate results in some circumstances. VDOT: Lessons Learned VDOT’s use of geophysical measurements has been driven to a large degree by karst issues the agency has faced in many areas of Virginia. Electrical and EM methods have been VDOT’s primary geophysical methods, though seismic methods such as refraction and surface wave methods have also been used on occasion. With the exception of GPR, VDOT contracts out all of its geophysical work and has found few problems with this approach. The agency’s experi- ence with geophysics has been generally positive, though some cases have had geophysical results that were ambiguous or inaccurate and that did not add value to VDOT projects. The case example presents a project where geophysical methods underperformed expectations for detecting the depth to bedrock. These results illustrate that the effectiveness of geophysical methods can vary depending on the problem investigated and the field conditions encountered. They also underscore the importance of performing ground truth measurements to corroborate the geophysical results.

44 Advancements in Use of Geophysical Methods for Transportation Projects Ohio Department of Transportation The synthesis survey revealed that the Ohio Department of Transportation (ODOT) applies geophysical methods on projects about six to 10 times per year, which the agency reported is more frequent than 5 years ago. ODOT reported experience using 13 of the 36 geophysical methods listed in the survey, for seven of the 11 applications listed in the survey. Detection of voids from abandoned mines and characterization of karst features are common applications for ODOT, as described in the following case examples. ODOT’s experiences with geophysical methods, including agency practices, commonly used methods, typical applications, and three example projects, are discussed in the following sections. History of Geophysics Use and Agency Practice ODOT has been using geophysical methods for approximately 15 years. ODOT initially con- tracted out geophysical work, but in the past 10 years it has developed in-house capabilities to perform ER and some seismic work. ODOT also has in-house capabilities to perform CPT, as well as drilling and sampling. The decision to use geophysical methods on specific projects typically originates in ODOT’s Field Exploration Group. The design engineer puts together an exploration request, which is (a) (b) Figure 31. Rock core from (a) borehole 14BH-001 starting at 14.5-ft depth and (b) 14BH-002 starting at 21-ft depth showing CH material to 35 ft (courtesy of VDOT).

Case Examples 45 further developed by the Field Exploration Group. When the group sees value in using geophysics for a project, it will suggest doing so in the field exploration program. ODOT reported that, as designers have become more familiar with the capabilities of the geophysical methods, they will sometimes request measurements directly. ODOT also reported that the in-house capability to perform geophysical measurements has made the agency more inclined to use equipment on marginal applications because it costs less and is relatively easy to deploy. Though ODOT has not quantified the cost savings from its use of geophysics, the agency has generally found that it is cost-effective to be able to perform geophysical measurements in-house. ODOT also reported cases where geophysics has not resulted in cost savings, particularly where the geophysical results were ambiguous and the site needed an extensive drilling and sampling program. Two individuals in ODOT’s Field Exploration Group have the experience to lead in-house geophysical investigations, and others are recruited to help in the field on specific projects. ODOT typically uses a three-man crew when both ER and seismic work is performed, and a two-man crew when only one method is used. Generally, training on the use of ER and seismic methods has come from vendors, and ODOT personnel have been satisfied with the level of training they have received. They see a need for a concise training course that is specific to geotechnical work as well as short tutorial videos on the various methods. In addition to their in-house capabilities, ODOT personnel also contract out work that they are not capable of performing. For example, ODOT has contracted out for FWI of seismic measurements to characterize underground mines, as described in one of the following case examples. Commonly Used Methods and Applications The synthesis survey results indicate that ODOT has used a variety of seismic and electrical methods, as well as GPR, gravity, and some borehole methods. In a phone interview, ODOT personnel reported that ER is the most commonly used method. They developed in-house capa- bilities to perform ER in 2010 and have used it for a wide range of applications. Initially, ODOT’s primary application for ER was identification of underground mines, karst features, and top of rock. Lately, the agency has had positive experiences using ER to determine the extent of shallow embankment failures and to delineate the extent of high-organic soil deposits. ODOT reported using its ER system about 3 or 4 times a year. ODOT also makes use of seismic methods, including refraction and surface wave methods. In the past year, it has developed in-house capabilities to perform ReMi measurements but has not used ReMi as extensively as ER. ODOT personnel collect both ReMi and ER data for some applications, such as development of rock profiles for rock-socketed drilled shafts, and have found ReMi to be very useful as an alternative to ER in urban environments where utilities may interfere with ER. During the phone interview, ODOT personnel emphasized the need for users to understand the capabilities and limitations of the methods. They noted that anomalies in geophysical measure- ments can be caused by multiple factors and that identifying a definitive interpretation is usually not possible. For this reason, ODOT personnel typically do not rely solely on geophysical measure- ments, except in cases of small, low-risk projects—such as characterizing shallow embankment failures—where misinterpretation of the results would not have significant consequences. ODOT Example Project 1: Irregular Bedrock at Interstate 70— Harper Road Crossing in Columbus, Ohio Interstate 70 crosses over a surface street, Harper Road, on the western side of Columbus, Ohio. The current three-span structure was constructed in 1973, with the geotechnical exploration

46 Advancements in Use of Geophysical Methods for Transportation Projects completed in 1967. In 2019, ODOT planned a project to replace the existing superstructure and widen the bridge. The existing structure is supported by driven H-piles bearing on limestone bedrock at the abutments and spread footings bearing on limestone bedrock at the piers. For the widening, a limited geotechnical exploration consisting of two borings to confirm top of bedrock was planned because of high confidence in the historical geotechnical information. One boring was planned at the rear pier and one boring at the forward abutment, with top of rock anticipated at around elevation 730 ft, or 7 ft deep at the rear pier and 25 ft deep at the forward abutment. During the early summer of 2017, both borings were completed. The first boring was completed within the footprint of the proposed rear pier, encountering top of bedrock at elevation 689.0 ft, or 48.5 ft below ground surface. The second boring was completed within the footprint of the proposed forward abutment and encountered bedrock at elevation 720.3 ft, or 34.5 ft below ground surface. The bedrock elevation at the rear pier was approximately 30 ft lower than anticipated. The structure is also located in an area known to have clay-filled voids associated with paleo-karst features. To better define the bedrock elevation, OODT completed ER surveys along the proposed alignments of both pier widenings. The results of the ER surveys indicated that the bedrock surface was highly variable and dropped quickly across the proposed area of the pier widening—not only at the rear pier but also at the forward pier location, as shown in Figure 32. The surveys were not long enough to completely image the depth of the bedrock surface. A third boring was planned at the forward pier location to confirm top of bedrock and encountered top of bedrock at elevation 699.8 ft, or 38.6 ft below ground surface. However, several large limestone boulders were encountered within the overburden, and a clay-filled void was encountered within the bedrock between elevations of 695.4 ft and 691.7 ft. Informed by the exploration results, ODOT personnel recommended that the proposed widened piers and abutments be supported on H-piles driven to refusal on bedrock. Construc- tion was started in the spring of 2019. Pile driving encountered a highly variable bedrock surface. The abutment piles were driven to 137% to 220% of the planned lengths at the rear abutment and 34% to 125% of the planned lengths at the forward abutment. At the rear pier, the piles were driven to 38% to 200%+ of the planned lengths. One pile at the rear pier was driven to 100 ft and stopped despite not bearing on bedrock. Before driving the forward pier piles, the area of the pile cap was excavated. Limestone bed- rock was observed at the bottom of the excavation, except for the last few feet furthest from the existing bridge where it appeared that the top of bedrock dropped off to greater depths. Consequently, only two piles were driven at the forward pier with the remainder of the foun- dation consisting of a spread footing bearing on bedrock. These piles were driven to 68% and 80% of the plan lengths. Through the use of ER surveys and a conventional drilling exploration, ODOT was able to anticipate a widely variable bedrock surface and select appropriate foundations. However, because of the limited geophysical program, the full breath of variability was not established. ODOT Example Project 2: Imaging a Shallow Embankment Failure in Northwest Ohio Paulding County in northwestern Ohio lies within the Paulding Clay Basin physiographic region, an area characterized as a nearly flat lacustrine plain. Soils are typically high-plastic clays, which tend to have poor long-term strength when used for embankment construction. The proj- ect area is where US-24 passes over a county road and a railroad spur. The embankments were

Case Examples 47 experiencing instability at all four quadrants of the overpass, with the northeastern quadrant exhibiting the greatest distress. The surface features indicated a shallow surficial sloughing of the outer embankment soils. A subsurface exploration was planned to determine the failure mode of the embankment using traditional borings, dynamic cone penetration (DCP) soundings, and geophysical surveys. Borings were completed at the top, mid-slope, and base of the embankment. Inclinometers were installed mid-slope and at the toe to determine a failure surface. The DCP soundings were completed in section with the borings to confirm the potential sliding surface. In addition to the traditional exploration techniques, an ER imaging survey was completed perpendicular to the roadway down the embankment slope. The traditional exploration and monitoring techniques indicated that the shallow embank- ment failure was a result of saturated and low-strength soils along the outer embankment slope. The ER survey indicated a shallow layer of higher-resistivity material underlain by low-resistivity material (Figure 33). This contrast in resistivity was probably attributable to higher moisture contents along the failure surface. The results from the ER survey showed strong agreement with slope inclinometer data from the site, which showed the slide surface at a depth of 4 ft to 6 ft. Figure 32. Subsurface profiles determined from ER measurements at rear pier (top) and forward pier (bottom) showing highly irregular bedrock surface (courtesy of ODOT).

48 Advancements in Use of Geophysical Methods for Transportation Projects Informed by the results of this project, ODOT expects to use ER surveys in the future to image shallow embankment failures and to minimize the disturbance and effort needed for traditional drilling exploration at mid-slope locations. ODOT Example Project 3: Full Waveform Inversion to Characterize Abandoned Mines While many other seismic methods are based on matching first arrival times of recorded waveforms, full waveform inversion (FWI) works by developing a subsurface model that provides a match to the full recorded waveform at each location. The 2D FWI technique was applied by researchers from Clarkson University to image abandoned underground coal mines under US 33 in Athens County, Ohio. The description that follows is summarized from Sullivan et al. (2016). The test area selected was thought to be a likely location of abandoned mines. Previous borings performed by ODOT showed a 1.5- to 2.5-m thick coal seam located about 12 m to 18 m below the surface, with the overburden consisting of clay shales and sandstones. A total length of 576 m was investigated using test segments of 36 m consisting of 24 4.5-Hz geophone receiv- ers spaced at intervals of 1.5 m (Figure 34). A sledgehammer source was used to excite energy at 25 locations spaced 1.5 m apart along the geophone spread, and a land streamer was used to collect the data, which allowed for rapid data collection along the roadway. The results indicated the presence of two anomalies along the profile. The results from one of the segments that contained an anomaly are shown in Figure 35. A low shear wave velocity anomaly is observed at a depth of about 15 m in the image. Borings performed at this location about 3 weeks after the measurement showed the presence of a void over the depth range of 13.8 m to 14.6 m. The other suspected void was also confirmed with drilling. This case example shows the capabilities of one of the more advanced seismic methods, full waveform inversion. The method successfully detected small voids at depths of about 15 m, although the size of the void appeared to be overestimated by the FWI results. Figure 33. Profile of ER measurements performed over shallow landslide with approximate slide surface shown with solid black line separating high-resistivity material from lower-resistivity material below (courtesy of ODOT).

Case Examples 49 ODOT: Lessons Learned ODOT has used geophysical measurements for a variety of applications, with many driven by issues associated with karst terrain (e.g., voids, irregular bedrock) and effects of abandoned mines on roadways. ODOT has also used geophysics to characterize organic deposits and the disturbed zone in shallow landslides. The agency has found it cost-effective to develop in-house capabilities for performing ER and some seismic measurements. Agency personnel have received training on the implementation of the methods primarily from equipment vendors and have reported good experiences. ODOT also reported that in-house capabilities make more frequent use of certain geophysical methods possible, and for more marginal applications, as well. ODOT contracts out for other geophysical measurements and reported good experiences with the FWI approach for detecting abandoned mines. Like the other agencies, ODOT emphasized the need for training resources to educate engineers on the capabilities and limitations of geophysical methods for engineering applications. New Jersey Department of Transportation According to the synthesis survey results, the New Jersey Department of Transportation (NJDOT) showed moderately frequent use of geophysical methods (three to five times per year) and a large number of geophysical methods used (12 of the 36 methods listed in the survey). The agency reported an increase in the use of geophysics over the past 5 years, which has been driven to a large extent by rock mechanics applications. NJDOT’s experiences with geophysical methods, including agency practices, commonly used methods, typical applications, and three example projects, are presented in the following sections. Figure 34. Land streamer of geophone used to collect data (Sullivan et al. 2016). Figure 35. Inverted image of shear wave velocities showing low-velocity region at depth of about 15 m (Sullivan et al. 2016).

50 Advancements in Use of Geophysical Methods for Transportation Projects History of Geophysics Use and Agency Practice Though NJDOT has used geophysical methods for approximately 20 years, it reported a slow build-up to its present level of use. NJDOT’s earliest use of geophysics involved sporadic use of GPR until about 20 years ago, when the agency began to apply other geophysical methods to the problem of abandoned mine characterization. At the time, the results were met with some skepti- cism, but over the past 20 years NJDOT has observed a cultural shift in the value of geophysical investigation and agencies’ openness to using geophysical methods on major projects. The Engi- neering Geology section of the NJDOT is largely responsible for suggesting, selecting, and con- tracting geophysical work. All of NJDOT’s geophysical work is contracted to external providers, and agency personnel reported few problems with the contracting process. They also noted how important it is for those involved in the contracting process to understand who is responsible for interpreting the data and how the data will ultimately be factored into the project design. NJDOT reported having contracted with geotechnical firms that are unfamiliar with the geo- physical methods the agency would like to use on a project. From their experience, NJDOT per- sonnel see a need for training resources for geotechnical engineers who are not well versed in the capabilities and limitations of geophysical methods, including appropriate applications, basics of data collection (such as equipment used and space requirements), and environmental constraints on applying the methods (e.g., cultural noise, interference). Because NJDOT personnel do not perform the measurements in-house, they do not have a need for detailed training on how to perform and interpret the measurements. Commonly Used Methods and Applications NJDOT has seen a dramatic increase in its use of geophysics over the past 5 years. This increase has been driven largely by rock mechanics applications. The agency has found geophysics particularly useful for rock mass characterization. The two most common methods used by the agency are seismic refraction, primarily for rippability evaluation, and borehole televiewers (acoustic and optical) for assessing rock characteristics such as weathering. Agency personnel have experience on projects where seismic velocities (and hence rippability) have varied greatly across the site and where the use of seismic methods, particularly MASW and refraction, to characterize the rock has helped avoid large cost overruns. Abandoned iron ore mines in the northern portion of New Jersey are often an issue for NJDOT. A common and early application of geophysics by NJDOT was to characterize such abandoned mines beneath roadways. For example, about 20 years ago, problems on Interstate 80 were caused by the collapse of abandoned mines beneath the roadway. Seismic tomography was used to image the site and found many low-velocity areas that indicated the possible pres- ence of voids. Ground truth boreholes drilled at the site found that many of the features were smaller in-filled voids that did not present a hazard. This experience underscored both the value of geophysics for identifying possible hazards and the need to tie the geophysical results to hard data, a point that was made by several of the agencies interviewed. NJDOT has also found value in using multiple methods to verify and calibrate the results against each other. One of the major practical issues the agency faces is the fact that most of its work is on high-capacity roadways, which necessitates performing the measurements at night to avoid noisy conditions. Even at night, the high-capacity roads are heavily traveled and preclude the use of some geophysical methods because of the noisy conditions. NJDOT Example Project 1: Determining the Depth of Bedrock beneath a Large-Block Talus Deposit This project was conducted along a busy (55,000 vehicles per day) interstate highway and involved a series of exposed near-vertical highway rock cuts interspersed with areas of soil slopes

Case Examples 51 overlying shallow bedrock. In addition, a unique aspect of the project site was an extensive (approximately 600,000-square-ft) talus field composed of large (on average, 15-cubic-ft) rock blocks overlying deeper bedrock. Rock types consisted of several members of a sedimentary suite. Preliminary alternative analyses required comprehensive site characterization and sub- surface exploration to determine top of rock, soil-bedrock, and talus-bedrock interface, as well as prevailing rock structure and rock properties. Control borings were performed in accessible areas (i.e., in the roadway and the lower soil slope areas). In addition, several borings were taken within the talus field. These talus borings were extremely labor intensive and costly; thus, a limited number were done. Optical televiewers and acoustic televiewers were implemented in all boreholes. However, because many areas were inaccessible to conventional boring equipment, additional geophysical methods were used to supplement the ground truth data of the borings. In areas where soil was present, seismic refraction was used to better define the soil-bedrock interface. Because the location’s talus field contains a large number of air voids and a corresponding lack of interstitial material, seismic refraction was eliminated as an applicable method. Alter- natively, horizontal/vertical inversion—specifically, single-station passive seismic stratigraphy (S-SPSS)—was used in this area. The S-SPSS used a surface instrument placed upon the talus at specific grid points to obtain point data. Ambient vibrations were measured to stitch together a shear wave velocity profile of the soil column. After data collection, the corresponding profile was used to assess the depth of hard, resonating layers such as bedrock. The S-SPSS results were encouraging, resulting in an improved subsurface soil-rock profile to assist in further evaluation of design alternatives, as well as constructability concerns. Interest- ingly, though reduction or elimination of background noise (such as highway traffic) is desirable in many geophysical applications, because S-SPSS uses ambient vibrations, it can be reasonably assumed that the high traffic volumes experienced within the project area likely resulted in more accurate results. However, as with all geophysical methods, there are limitations to this technol- ogy, and it is best used in conjunction with other applications where suitable. NJDOT Example Project 2: Reconfiguration and Expansion of an Existing Interchange This project involved the planned reconfiguration and expansion of an existing interchange of an interstate highway and state highway. In addition to numerous proposed new structures, new ramps, and the widening or reconstruction of existing ramps, six known abandoned iron ore mines from the early 1800s are located within the project limits. Preliminary alternatives analyses required comprehensive subsurface exploration and site characterization to determine the soil profile, top-of-rock interface, and prevailing rock struc- ture and rock properties. These investigations also helped assess whether there was a need to incorporate mine remediation measures into the project design. The subsurface investigation was conducted in three separate phases. The first phase mapped the known abandoned mines in relation to the project area and identified potential issues; the remaining two phases of investiga- tion were developed for the furtherance of the project design. During Phase One, MASW and 2D ERI were chosen to map the known mines and compare the size, location, and current conditions against the existing literature and maps of the mines themselves. The operations were conducted overnight and used lane closures in an effort to reduce background noise caused by typical daytime traffic volume. Phase Two consisted of seismic refraction surveys along a proposed ramp alignment to evaluate rippability and delineate top of rock. The results indicated seismic velocities for the rock mass on the ramp alignment approaching 40,000 ft per second. In addition, the results provided corroborating evidence of previous mine remediation on adjoining properties adjacent to NJDOT right-of-way.

52 Advancements in Use of Geophysical Methods for Transportation Projects Phase Three consisted of a series of borings, including rock coring, for a proposed widen- ing of one of the existing ramps. To facilitate structural rock analysis and rock cut design along the ramp, borehole logging with optical televiewers was performed on all borings. Interference from magnetite veins within the rock mass made it necessary to adjust the boring angle. Seismic refraction was also performed in this area; results showed lower seismic velocities than measured elsewhere on the project. The overall results of the three geophysical investigations supplemented the characterization of the site conditions. The staggered approach of the individual phases afforded NJDOT the ability to tailor each successive phase according to the results of the preceding phase. Every geophysical appli- cation used had its inherent advantages and limitations, and bundling several methods with borings to provide ground truth was preferable to using a single method. In this way, the project developed a more comprehensive model of the existing subsurface conditions and site characteristics. NJDOT Example Project 3: Costly Project before the Use of Geophysics NJDOT also discussed a case example from more than 20 ago—before geophysical methods were used routinely by NJDOT. The project is a good example of the costs associated with not using geophysics. The project involved excavation in rock in the karst terrain of western New Jersey, where ground conditions consisted of pinnacled bedrock. The soil borings performed at the site picked up the highest points in the pinnacled rock such that the bedrock surface was interpreted to be at a fairly consistent depth and rather shallow. When the cut was performed, the bedrock depth was found to be highly erratic and much deeper in several locations. Although the presence of soil allowed for easier excavation than expected, the absence of rock required re-engineering of all of the slopes and the construction of retaining walls. The resulting multimillion-dollar claim could potentially have been avoided if geophysics had been among the tools used by NJDOT at that time. NJDOT reported that if the same project were performed today, geophysics would be used and the likelihood of this problem and the resulting claim would be significantly reduced. NJDOT: Lessons Learned NJDOT has used geophysics for about 20 years but reported a more rapid expansion of use in the past 5 years. In the past 20 years, the agency has observed a culture shift as engineers have become more open to using geophysics on their projects. NJDOT has found the most value in applying geophysics to rock problems, particularly seismic methods that can be used to characterize rock rippability and bedrock depth. NJDOT’s cost savings have been primarily realized through improved subsurface characterization, which has helped avoid claims. The agency has also had cases where geophysics was the only viable option to obtain the information needed, as was the case for the talus deposit case example presented earlier. All of NJDOT’s geophysical work is contracted out. The agency has experienced some frustration over dealing with geotechnical engineering firms that are unfamiliar with the geophysical methods the agency would like to use. Agency personnel see a need for better training of engineers on the capabilities and limitations of geophysical methods. California Department of Transportation The California Department of Transportation (Caltrans) indicated the highest level of use of geophysics among the agencies that responded to the survey. Caltrans reported using 20 of the 36 methods listed in the survey, and applied those methods to all 11 of the applications listed in the survey. Like many of the other agencies, Caltrans reported more frequent use of geophysics

Case Examples 53 now compared with 5 years ago. Caltrans was also the only agency interviewed that had a formal geophysics group within the agency. Caltrans’s experiences with geophysical methods, including agency practices, commonly used methods, typical applications, and two example projects, are presented in the following sections. History of Geophysics Use and Agency Practice The use of geophysics at Caltrans dates back to the 1950s, as evidenced by photographs from that period showing refraction measurements performed on a bridge project. In the 1970s, several research projects in California were initiated to investigate the correlation between p-wave velocity and excavation potential. Until the 1990s, geophysical work at Caltrans consisted primarily of refraction measurements for determining rock rippability and depth of bedrock. After the Loma Prieta earthquake in 1989, a large effort was undertaken to seismically retrofit toll bridges in California, which required p-wave and s-wave velocity logging. In 1993, Caltrans acquired a P-S suspension log, which was used extensively in onshore and especially offshore applications. Caltrans continued to develop additional borehole logging capabilities, including conductivity measurements and acoustic and optical televiewers. In 1998, a formal group dedi- cated to geophysics was created. The group initially consisted of two people and currently consists of eight individuals, including geologists, technicians, and engineering geophysicists. The mission of the Geophysics and Geology Branch is “to provide geologic and engineering data through nondestructive geophysical methods in support of foundation and geotechnical studies for state highway construction projects.” The group currently has extensive in-house capabilities to perform surface surveys, including seismic refraction, refraction tomography, ReMi, GPR, EM induction, resistivity, and magnetometry, as well as borehole logging methods, including PS logging, borehole caliper, natural gamma, induction, acoustic televiewer, full waveform sonic, and resistivity. In addition, oversight and review of contracted geophysical projects are conducted by the group. Caltrans reported that surface wave measurements and seismic reflection are typically contracted out, and the agency reported few problems with the contracting process. Policies for the use of geophysics are presented in the Caltrans (2019) Geotechnical Manual within the Geotechnical Investigations section. The manual states that geophysical methods are not typically used in the project planning phase unless the project needs justify such methods. After the project comes into the design office, the Geophysics and Geology Branch acts as a support office to bolster logging of boreholes or augment drilling information with surface geophysical methods. Cost savings are realized both through using the cheaper geophysical measurement in lieu of additional borings and—more significantly—through avoiding the unforeseen costs of unexpected site conditions. Another role of the Geophysics and Geology Branch is to train and inform engineers about the capabilities of geophysical methods. The branch has found that, because of staff turnover, many engineers coming on board are not well versed in how geophysical techniques can be used on their projects. The branch thus sees a need for training materials that would help educate geotechnical engineers on the capabilities of modern geophysical methods, such as a National Highway Institute course on geophysics. Commonly Used Methods and Applications Caltrans has extensive in-house capabilities to perform both surface and borehole geophysical measurements. It has a borehole logging truck that travels throughout the state with the capa- bility to perform PS logging, borehole caliper, natural gamma, induction, acoustic televiewer, full waveform sonic, and resistivity measurements. The borehole measurements are used to

54 Advancements in Use of Geophysical Methods for Transportation Projects provide a continuous record of the physical properties of the soil and rock; this augments the discrete information obtained from lab analysis. Caltrans reported extensive use of the acoustic televiewer in agency projects that seek to find orientations of failure surfaces, fractures, and bedding. Acoustic logging negates the need to perform other more cumbersome methods, such as oriented core or manned-hole logging. An example of an acoustic televiewer log from a foundation investigation is shown in Figure 36. Caltrans frequently uses surface geophysical methods to interpolate geologic data into areas that have not been explored with boreholes. The surface methods employed by the agency most routinely are p-wave refraction tomography and GPR. An example of p-wave tomography and of GPR results are presented in the sections that follow. Caltrans has a number of GPR units that are used for both geophysical and nondestructive evaluation applications. Among these units are a 3D radar unit that is capable of mapping beneath roadways at high speeds. In addition to seismic tomography and GPR, Caltrans reported frequent use of resistivity for mapping sand and clay deposits, as well as frequency-domain EM for soil conductivity mapping and time-domain EM for utility mapping. Caltrans Example Project 1: Freeway Improvement Project at Interstate 80 and Willow Avenue in Contra Costa County Refraction tomography is the most common surface method used by Caltrans and has largely replaced conventional refraction processing. An example of the use of borehole-to-surface Figure 36. Example of acoustic televiewer log performed by Caltrans from foundation investigation. Shows identification of open joint with aperture of about 10 cm (courtesy of Caltrans).

Case Examples 55 refraction tomography for a freeway improvement project at Interstate 80 and Willow Avenue in Contra Costa County is presented here. The site is located near a heavily traveled urban freeway, whose traffic produced significant broad-band seismic noise. Refraction tomography measurements were performed to fill in information between two boreholes spaced approximately 60 m apart at the site, as shown in Figure 37. The depth to rock at these two boreholes varied from about 3 m to the north to 20 m to the south. Because a limited footprint was available for performing surface refraction mea- surements, two source shots were performed at depth in the borehole in addition to the surface shots, in an effort to adequately image the deep end of the profile. The ray coverage produced by this shot arrangement is shown in Figure 38, where the hit count indicates the number of rays pass- ing through a pixel. The refraction tomography results provided an excellent image of the variable subsurface conditions between the boreholes, as shown in Figure 39. For comparison, the pseudo ray path model without the borehole shots is shown in Figure 40. When the borehole shots are not included, the depth of resolution is greatly limited and the measurement is unable to image the deeper rock. These results illustrate the dramatic effect of adding just a few borehole shots on the depth resolution of refraction tomography measurements. Caltrans Example Project 2: GPR to Determine Stratigraphy Caltrans provided information on a case example of using GPR to image stratigraphy. The project involved an area adjacent to an existing roadcut, referred to as Big Pumice Cut, along US 395. Ground-penetrating radar data were collected along transects of a southwest-facing natural hillslope at different distances from the edge of an existing cut, as shown in Figure 41. The purpose of the investigation was to determine if the stratigraphy exposed in the cut extended back into the hillslope. The cut exposes the Sherwin Till overlain by the Bishop Tuff, as shown in Figure 41. Ground-penetrating radar surveys were conducted using 50 MHz antennas to obtain reflec- tion data in the area under investigation. Estimates of the contact depth between the Bishop Tuff and Sherwin Till were based on extending the exposed stratigraphy back into the hillslope. Figure 37. Plan view of location of refraction tomography measurements used to fill in between two boreholes at project site (courtesy of Caltrans).

56 Advancements in Use of Geophysical Methods for Transportation Projects Figure 38. Pseudo ray path model for velocity section shown in Figure 39 (courtesy of Caltrans). Figure 39. Velocity model and lithology interpretation between boreholes using borehole logs and borehole-to-surface tomography at I-80 and Willow site (courtesy of Caltrans).

Case Examples 57 Figure 40. Pseudo ray path model using only surface sources, showing change in depth of investigation (courtesy of Caltrans). Figure 41. Road cut along US-395 showing locations of three transects (courtesy of Caltrans). The GPR measurements showed clear reflections that were interpreted to be the interface between the Bishop Tuff and Sherwin Till, as shown in Figure 42. Seismic refraction data of the Bishop Tuff and Sherwin Till showed that the materials could not be distinguished by seismic velocity, and resistivity measurements were unsuccessful because of the high resistance of the dry tuff. Caltrans: Lessons Learned Caltrans makes extensive use of geophysical methods and has a dedicated in-house geo- physics group with the capability of performing a wide range of borehole and surface geophysical

58 Advancements in Use of Geophysical Methods for Transportation Projects Figure 42. GPR reflections interpreted to represent contact between Sherwin Till and overlying Bishop Tuff (courtesy of Caltrans). techniques. Caltrans has found that selective use of geophysics can improve the quality of geo- technical investigations and produce significant cost savings. These cost savings are realized through the use of in-house expertise to perform routine geophysical services. The case examples presented illustrate results obtained from two of the agency’s most commonly used methods— seismic refraction tomography and GPR. The favorable results from geophysical investigations that Caltrans has experienced have demonstrated the effectiveness of geophysical service within the agency. The Caltrans geophysi- cal group also educates engineers in Caltrans on the capabilities and limitations of geophysical measurements. Like many of the other agencies interviewed, Caltrans sees a need for training resources to help educate engineers on geophysics and would make use of such resources if they were developed. Lessons Learned from All Case Examples Although the experiences of the five agencies described in this chapter vary significantly, they also overlap in several areas. Each interview provided valuable lessons on how geophysical capa- bilities have been developed, how they are typically used, and benefits to their use in transporta- tion projects. A summary of lessons learned from these five agencies is as follows. • All the agencies interviewed agreed that geophysical methods act as a cost savings tool when they are applied to the appropriate problem and field conditions. The degree of cost savings

Case Examples 59 varies by project and has not been quantified by the agencies in most cases. The main source of cost savings mentioned by most of the agencies was avoidance of claims or unanticipated delays through better characterization of ground conditions. • Implementation of geophysics differs significantly among the agencies interviewed. Three of the agencies, MnDOT, ODOT, and Caltrans, have developed in-house capabilities to perform at least some of the more routine geophysical methods. Caltrans in particular stands out as having a broad range of in-house capabilities and a group dedicated to geophysics. These three agencies all credited in-house capabilities for both cost savings in performing the measure- ments and an increased likelihood of using the methods. These agencies also contract out some geophysical work. Two of the agencies, VDOT and NJDOT, obtain all or nearly all geo- physical measurements through external contracts. None of the agencies reported significant problems with the contracting process. • As expected, given the geographic distribution of the agencies interviewed, the applications and primary methods used differed among the agencies. For MnDOT, VDOT, and ODOT, identification of voids and cavities in karst conditions and of abandoned mines was a com- mon application. These agencies also identified ER as the most common method used. For NJDOT and Caltrans, seismic methods, particularly refraction, seismic tomography, and GPR, were the most common methods used. NJDOT used geophysics primarily for rock- related issues, whereas Caltrans used it for a wide range of soil and rock applications. • Some agencies reported experiences where geophysical measurements provided ambiguous, confusing, or inaccurate results. Agency personnel emphasized the need to apply the methods under the right conditions and to avoid using certain methods at sites with known sources of interference. • All the agencies emphasized the need to correlate the results with ground truth from borehole data. Many cautioned against relying too much on geophysical measurements because of the uncertainty in relating geophysical anomalies to specific subsurface features. • Personnel from two of the agencies that performed geophysical measurements in-house— MnDOT and ODOT—developed the technical skills to perform the measurements through vendor training. Personnel from each agency expressed satisfaction with the training and level of support they had received from the vendor. The Caltrans group hires geophysicists and geologists with the necessary background and does not rely as much on vendor training. • Most of the agencies mentioned experiences working with geotechnical engineers or contrac- tors who were unfamiliar with geophysical methods. Personnel from each of the agencies thought there was a need for better training of geotechnical engineers on the capabilities and limitations of geophysical methods. • The extensive use of acoustic televiewers was another consistent comment from most of the agencies. • Many of the agencies indicated a culture change over the past decade or so, during which engineers have become more comfortable with using geophysical methods.