Managing Geotechnical Risks in Design–Build Projects (2018)

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.

49 Chapter 3: Findings and Applications 3.1 Introduction The completed research developed a number of findings as a result of the research instruments listed in Chapter 2. These will be presented in this chapter by research instrument. 3.2 Summary and Analysis of the Literature Review and Content Analysis. The literature was comprehensively covered in NCHRP Synthesis 429 up to its 2011 publication date. Therefore, the research team merely needed to update it to account for new developments since that time. Two additional bodies of information were analyzed: geotechnical risk and alternative project delivery with a focus on DB. 3.2.1 Geotechnical Risk Literature NCHRP Synthesis 484: Influence of Geotechnical Investigation and Subsurface Conditions on Claims, Change Orders, and Overruns, (Boeckmann and Loehr 2016) provided the broad coverage of project types and was specific to highway construction. Therefore, its findings were deemed to be valuable to the NCHRP 24-44 project. NCHRP Synthesis 484 had several findings that directly relate to the DB context of geotechnical risk management. They are summarized as follows:  The majority of DOTs have published minimum subsurface investigations standards that comply with those provided in the AASHTO guide specifications.  “The total share of claims, change orders, and cost overruns attributed to subsurface conditions out of all claims, change orders, and cost overruns was 5% by number and 7% by cost” (Boeckmann and Loehr 2016).

50  “The cost of subsurface condition change orders approaches 1% of the agencies’ total budgets for new construction” (Boeckmann and Loehr 2016).  Increasing the amount of subsurface investigation, while beneficial to the design process, had no impact on the claims, change orders, and cost overruns.  “Implementing minimum standards for subsurface investigation and site characterization was reported to reduce claims, change orders, and cost overruns” (Boeckmann and Loehr 2016).  Increased communication and training on the topic “appear to be a critical component to realizing the benefits of improvements to site characterization practices” (Boeckmann and Loehr 2016). A paper that analyzed the “misallocation of risk” on highway projects (Hanna et al. 2015) found that five of the top nine misallocated risks were associated with subsurface conditions. Those risks are: inadequate geotechnical investigations, differing site conditions, unstable subgrade material, unforeseen utilities, and relocations due to utilities. All of these risks apply in the DB context, and the paper reinforces the findings of the two NCHRP syntheses with regard to the need to apply a high standard of care in the allocation of risk in DB contracts. Figure 3.1 comes from the paper and provides a succinct methodology for evaluating the potential for differing site conditions claims. The figure’s logic can be used to determine where the subsurface risk can best be allocated based on answering the questions posed in the flow chart.

51 Figure 3.1 Differing Site Conditions Risk Liability Flow Chart (Hanna et al. 2015). Another paper contributed a tool for developing a sequence of construction that is cognizant of geotechnical risk (Tapia et al. 2017). The approach comes out of the recently completed Panama Canal DB project where linear scheduling was used on major features of work that involved significant amounts geotechnical engineering and construction. The example, the Borinquen Dam 1E, used in the paper was a 2.3-kilometer long rockfill dam with a residual soil impervious core and several zones for filters and blankets. It featured extensive foundation treatment works and a 16-meter deep grout curtain. The approach involved first converting the subsurface geotechnical conditions expressed in the project’s geotechnical baseline report (GBR) into color-coded risk category zones as shown in Figure 3.2.

52 Figure 3.2 Color Coding for a Section of the Borinquen Dam 1E Foundation Geological Profile (Tapia et al. 2017). Figure 3.3 is an actual as-built linear schedule from the Borinquen Dam projects where the geotechnical risk profile is shown on the bottom of the linear schedule. Without getting into the details of linear scheduling, the figure shows that the contractor’s approved baseline schedule showed it planned to start grouting at either end of the project and work towards the middle. After evaluating the construction sequence in light of the geotechnical risk, a decision was made to reverse the process as shown in the solid blue line and start grouting in the zone of highest geotechnical risk and work from the middle to both sides. The time scale is on the y-axis and the height of the blue-colored areas indicates how long it took to actually complete the grout curtain. Thus, this tool provides a means to relate the geotechnical risk visually to the planned construction means and methods. The decision to reverse the directions of the grout curtain crews in this DB project reinforces the major finding of NCHRP Synthesis 429 that the best way to manage DB

53 geotechnical risk is to start the subsurface work (in this case the grout curtain) as soon as practical. It must be noted that the design-builder was compensated for the differing site conditions shown in the red zones. Figure 3.3 Example of DB Decision Change Based on Geotechnical Risk Assessment (Tapia et al. 2017). This particular tool holds promise for use in other types of DB projects. Figure 3.3 is an example of construction visualization that permits the transfer of detailed geotechnical information in a manner that can be easily understood by DB project participants that do not have a specialty in geotechnical engineering. On a slightly different level, since DSC claims generally involve quantifying the delays experienced after the DSC was identified, the production-based schedule produced using linear scheduling also provides a direct check on the contractor’s progress prior to

54 the DSC delay. The analyst can compare the pre-DSC production rate with the planned production rates in the approved baseline schedule and quickly determine whether the contractor was achieving its as-bid production and whether it was behind schedule at the time the DSC was discovered (McLain et al. 2014; Lopez del Puerto and Gransberg 2008). 3.2.2 Design-Build Literature The major finding in the DB literature was the advent of the use of progressive DB (PDB) to the suite of project delivery methods available to state DOTs. PDB is organized in much the same structure as construction manager/general contractor (CMGC), except that the construction contractor and design consultant have contract privity. The PDB procurement can be either a qualifications-based selection of the design-builder followed by a negotiation of a guaranteed maximum price (GMP) during design development, or a “qualifications-driven” approach where the design-builder is asked to quote its fees for preconstruction, design, and construction management are included as a lump sum with the remainder of the construction cost being a negotiated GMP (MSHA 2016). The Maryland State Highway Administration (MSHA) was the first state DOT to experiment with the process and a case study of that project is contained in this report. However, PDB is not new to the transportation industry having been used successfully for years in the airport and rail transit sectors (Touran et al. 2010). PDB’s attraction in the geotechnical risk management realm is the ability to negotiate the geotechnical risk after award and after the geotechnical investigation has been completed rather than depending on the DSC clause to allocate the subsurface risk. With PDB, the first design and construction package released for construction could be to commence the subsurface investigation and conduct selected excavation on the project site to identify where the geotechnical issues will be encountered and their magnitude. Thus, a fair and equitable amount for the realized subsurface

55 risk can be established. On the owner’s side, PDB provides a means to avoid paying for unrealized contingencies in the design-builder’s lump sum price. On the design-builder’s side, since the contract has been awarded, there is no longer a possibility that over-pricing the geotechnical risk will cause a loss of the project to a less sophisticated or less well-informed competitor. According to Loulakis et al. (2016), the states of Arizona, Colorado, Florida, Maryland Nebraska, North Carolina, and Virginia have legislation that would permit them to use PDB if it was desired. The same source lists the following “reasons” owners choose PDB over lump sum DB:  “Ability to introduce design-builder to the project as early as possible  Design-builder as strategic partner  Enhanced risk transfer – avoidance of Spearin liability  Facilitates having design-builder involved in permit and other development activity  Substantially reduced cost and time for procurement helps everyone, including small design firms  Owner interest in being involved in design and procurement  Owner access to real-time [pricing] information  Open book pricing allows transparency into design-builder’s costs  Process fosters collaborative relationships” (Loulakis et al. 2016). 3.2.3 Findings from the Literature An analysis of the literature supports two findings and adds several possible tools for inclusion in the Guidelines developed by the NCHRP 24-44 research team. 1. As will be demonstrated in the survey findings, the decision to use DB project delivery is rarely, if ever, influenced by the perceived risk of geotechnical conditions on the project

56 site. Given that, the question is no longer whether DB is an appropriate procurement method, but rather how to mitigate the risks posed by geotechnical uncertainty at the time of DB contract award. The work completed by Boeckmann, and Loehr (2016) and Hanna et al. (2015) clearly supports the need for guidance on geotechnical risk management and underscores the importance of that guidance when the project’s delivery schedule is accelerated by the decision to use DB. 2. The literature strongly supports the notion that the DB procurement process should be planned in a manner the permits the early start of subsurface investigation, design and construction by the winning DB team. While not conclusive, the analysis appears to suggest that DB projects with high levels of geotechnical uncertainty would benefit from separating the analysis of geotechnical risk from the rest of the project. This in turn would raise the visibility of the geotechnical effort as well as permit short-term decisions regarding risk mitigation measures, such as additional investigation or the early start of site construction, to be made before the rest of the project’s risk profile is determined. 3. Tools like linear scheduling (Tapia et. al 2017), construction geotechnical risk visualization (McLain et al. 2014), and PDB (Loulakis et al. 2016; MSHA 2016) seem to offer potential approaches to mitigating geotechnical risk. 3.2.4 Findings from the Content Analysis This section summarizes the findings from the analysis of 59 different RFP’s across the United States regarding the way geotechnical risks are managed in DB projects at the time of advertising. Table 3.1 shows the list of projects that were analyzed, its contract value, and the type of document that was used.

57 Table 3.1 List of Projects in Content Analysis State Agency Project Value of project (thousands $) Type document Publish Year AK ADOT&PF 7 Mile ADA Accessibility 1,000 RFP 2011 AK Alaska RR Corporation Northern Rail Extension 100,000 RFP 2010 CA Riverside County State Route 91 Corridor Projects 1,700,000 RFP 2012 CA California High Speed Rail Authority California High Speed Train Project 1,800,000 RFP 2012 CA Caltrans Fresno 180 55,000 RFP 2011 CA Caltrans Los Angeles I-10/605 Interchange 61,800 RFP 2011 CA Caltrans Cajon Pass 115,000 RFP 2013 DE DelDOT Indian River Inlet Bridge 150,000 RFP 2008 DE DelDOT Dover Transit Center 5,000 RFP 2009 FL Florida DOT Central Florida Commuter Rail Transit 158,000 RFP 2008 FL Hillsborough County Tampa Airport 925 RFQ 2009 FL FlaDOT Hathaway Bridge 86,000 RFP 2000 FL FlaDOT Southbound I-95 Weigh Station 15,500 RFP 2008 FL FlaDOT I-75 Freeway Management System 7,000 RFP 2008 FL FlaDOT Central Florida Commuter Rail Transit 158,000 RFP 2008 GA GaDOT P.I. #'s 410240, 410500, 511084 33,000 RFQ - GA Corps of Engineers Aircraft Support Equip Paint Facility 5,000 RFP - ID Idaho Transportation Department SH-44 Linder Road to Ballantyne Lane 10,000 RFP 2012 IN Indiana Transportation Finance Authority I-90 Indiana Toll Road Fiber Optics 3,000 RFP 2000 KS Kansas DOT Johnson County Gateway 288,165 RFP 2013 KY KTC US 27 Widening & Relocation 12,000 LOI 2007 KY KTC Wilson Road 3,200 RFP 2006 ME MaineDOT Howland Piscataquis Bridge #3040 - RFP 2010 ME MaineDOT South Portland Vetran's Memorial Bridge #3945 - RFP 2009 ME MaineDOT Caribou Connector - RFP 2010 MD MDSHA InterCounty Connector 40,000 RFP 2006 MI MDOT M-222 Slope Stability Project 4,000 RFQ 2011 MN MnDOT TH 100 20,000 RFP 2001 MN MnDOT TH 52 200,000 RFQ 2002 MN MnDOT TH-14 10,000 RFP 2001 MS MissDOT I-59 Bridge Widening Project 10,000 RFP 2009

58 State Agency Project Value of project (thousands $) Type document Publish Year MO MoDOT "The New I-64" 535,000 RFP 2005 MT MDT US-2 Rockfall Mitigation 3,000 RFQ 2011 NV City of Reno ReTrac Reno Railway 157,000 RFP 2001 NV Nevada DOT Project NEON DB 550,000 RFP 2015 NM NMDOT I-40/Coors interchange 92,000 RFP 2006 NC NCDOT C2011288 Mecklenburg County - RFP - NC NCDOT Rec. I-85 from NC150 to I-85 Business 65,000 RFP 2010 ND NDDOT Highway 13 Box Culvert Replacement Project 600 RFQ - OH ODOT I-70/I-670 Interchange - RFP 2011 OR Private Owner Willamette Auto Dealership - RFP - SC SCDOT Cooper river Bridge 540,000 RFP 2001 SC SCDOT Port Access Road - RFP 2015 TN TNDOT SR-317 - RFP 2009 TX TxDOT Dallas Horseshoe 620,000 RFP 2012 UT UDOT I-15 Salt Lake City 1,590,000 RFP 1997 UT UDOT SR 201 Design-Build 43,000 RFP 2003 UT UDOT SR-265 University Parway at UVSC 1,400 RFP 2004 UT UDOT I-215 Tunnel - RFP 2004 VT Vtrans Richmond Checkerhouse Bridge 14,000 RFQ 2010 VA VDOT Rest and State Welcome Center - RFP 2002 VA VDOT Garden Creek Canal Bridge 1,100 RFP 2006 VA VDOT I-295 Interchange Meadowville Road 13,500 RFP 2010 VA VDOT 713 Bridge 600 RFP 2006 VA VDOT Buffalo Creek Bridge 2,900 RFP WA WSDOT SR520 Pontoon Bridge 600,000 RFP 2009 WA WSDOT SR 167 South Hot Lane Project 54,000 RFP 2014 WA WSDOT I-405 NE 6th to I-5 Widening andExpress Toll Lane 250,000 RFP 2006 WA WSDOT SR 9 / SR 92 Intersection 10,000 RFP 2011 The content analysis consisted of 18 geotechnical-related aspects that were thoroughly reviewed in each of the contractual documents shown in Table 3.1. This section details the results from analyzing these aspects in terms of the frequency in which they appeared. The first aspect analyzed was the amount of information included by the owner in the contract for proposers to use in their bids. Eight different ‘levels’ of geotechnical investigations

59 were identified by the research team and their frequency of occurrence in the analyzed documents is shown in Table 3.2. It was found that 24 out of 53 occurrences were at the level of geotechnical data report. This means that owners most often do not include any type of analysis of the subsurface exploration data in their DB contracts, limiting the studies to only provide raw data without interpretation or engineering behind it. Table 3.2 Content Analysis: Scope of Information Provided Pre-bid. Description Freq. Scope of information provided pre-bid No geotechnical data 0 Nonbinding geotechnical data - Reconnaissance Report (Review of records observations from site) 4 Geotechnical Data Report (Review of records and limited investigation data) 24 Geotechnical summary report (Review of records and Boring logs) 9 Preliminary geotechnical design report (Partial geotechnical investigation) 9 Final geotechnical design report - Full subsurface investigation for all structures and geotechnical features 2 Geotechnical baseline report (GBR) 5 Geotechnical Interpretive Report (GIR) 0 Knowing the type of information included in the contractual documents for each of the analyzed projects, the next aspect of interest was whether or not the owner allows interaction in the geotechnical information gathering process, i.e., if proposers are allowed to request additional borings or information. Out of the 59 projects, interaction was specifically not allowed in 19 projects. Considering that there were 14 instances where the analyzed documents did not allow for that determination, it was found that the majority of the projects did allow some level of interaction, as can be shown in Table 3.3 below. Table 3.3 Content Analysis: Extent to Which the Geotechnical Gathering Process Is Interactive Description Freq. Extent to which the geotechnical gathering process is interactive (e.g., Can proposers request supplemental borings?) Specific geotech requests allowed 15 Pre-bid RFIs could include geotech 11 No interaction 19 Can't tell 14

60 Another aspect analyzed was whether or not owners explicitly describe a project as one with significant geotechnical-related issues. By doing so, the owner might be able to obtain insight from the proposers as to identify ways to mitigate the risk, or avoid having proposers that did not consider a potential problem in their proposals. Table 3.4 shows the findings of this inquiry where it stands out that in a total of 14 projects, the owner did expressly identify the project as one with significant geotechnical issues in the contract, which makes this a practice to consider. Table 3.4 Content Analysis: Description of Project with Significant Geotechnical Issues Description Freq. Does scope describe project as one with significant geotechnical issues? yes 14 no 30 can't tell 13 One key characteristic of the DB project delivery method is that the design effort is performed after the contract is awarded. This creates the need for additional (or complete) geotechnical studies to be performed by the contractor to accompany the design. The owner may require the contractor to submit specific geotechnical reports after the contract is awarded. Four different types of documents were identified by the research team to be commonly required. Table 3.5 shows the results of this inquiry on the 59 selected projects. Table 3.5 Content Analysis: Information Required from Contractors in Post-award Description Freq. Post-award information required from contractors, including submittals. Final geotechnical investigation 35 Final geotechnical report 34 Geotechnical as schedule milestone 7 Geotechnical in quality plan 11 None found 18

61 Since the owner allocates design responsibilities to the contractor in DB delivery, contracts often include qualification requirements as a way to ensure the design is performed up to certain standards. The research team identified 11 different geotechnical-related aspects that owners can include in the evaluation as part of the contractor selection process. The results from analyzing the content from the selected contracts are shown in Table 3.6 below. Table 3.6 Content Analysis: Evaluation Criteria in Selection Process Description Freq. Selection process- evaluation of: Specific geotechnical qualifications 28 Specific geotechnical experience 25 Proof of local experience 21 Geotechnical designer - of record 19 Geotechnical narrative 10 Geotechnical in evaluation plan 14 Geotechnical risk narrative 8 Warranties scored - not specific to geotechnical but could be 5 Geotechnical warranties scored 0 Geotechnical in price proposal 8 References required from past project geotechnical 3 Another way to incorporate geotechnical considerations into the contractor selection process is to directly assign a weight for geotechnical factors in the evaluation of proposals. The content analysis of contractual documents showed that 41 out of 57 projects did not include any type of geotechnical considerations in the evaluation of proposals. However, 16 projects did incorporate geotechnical factors in their evaluations in varying degrees (Table 3.7). Table 3.7 Content Analysis: Weight of Geotechnical Factors in Proposal Evaluation Description Freq. Weight of geotechnical factors None 41 Minor <10% 12 Some 11-20% 1 Heavy >20% 3

62 A way in which owners seek to obtain information regarding the assumptions and interpretations that the contractor has made in order to submit a proposal is to require specific geotechnical-related supporting documents. The research team identified five different types of supporting information that are required by some owners in the requests for proposal, Table 3.8 shows the frequency in which these requirements were found in the 59 projects that conform the content analysis; 34 projects did not include any explicit requirement and varying instances of required design and performance criteria were found in 47 projects. Table 3.8 Content Analysis: Design and Performance Criteria to Be Submitted in Proposal Description Freq. Design and performance criteria to be submitted in proposal. Safety factors 6 Geotechnical design values 11 Preliminary design of geotechnical 12 Geotechnical design approach narrative 14 List of geotechnical assumptions 4 None found 34 Some of the analyzed contracts required the use of geotechnical-related performance verification as a way to document and verify the actual conditions met on site once the work starts; 52 instances were found where the owner required either test piling, a foundation certification package, or instrumentation to measure the performance of the subsurface. Table 3.9 shows the frequency in which these instances were found. Table 3.9 Content Analysis: Use of Performance Verification and Measurement Methods Description Freq. Use of performance verification and measurement methods Test piling 18 Foundation certification package 10 Instrumentation 24 As explained in other sections of this report, the differing site conditions clause is the contractual way to compensate the contractor for actual conditions on the site that materially differ

63 from those expected. There are many different wordings for this DSC clause in construction contracts. For instance, the geotechnical studies performed by the owner could be directly referenced in the clause as basis for determining the occurrence of a differing site condition or not; the referred document can be a geotechnical baseline or only the raw data; or there is not a DSC clause in the contract. Table 3.10 shows the findings from the 59 analyzed projects in the content analysis, it is important to note that a significant number of projects (24) where found to not include any differing site conditions clause, which is a significant finding that separated geotechnical risk management practices from those in DBB delivered projects. Table 3.10 Content Analysis: Differing Site Conditions Clause Description Freq Differing site conditions clause yes 35 no 24 includes reference to RFP geotechnical data 10 reference to RFP GBR 5 reference to post-award geotechnical reports 2 no specific geotechnical reference in clause 16 other 0 As a way to mitigate potential geotechnical-related failures in the operation phase of the project an owner might include a requirement for specific geotechnical-related performance warranties by the contractor. The performed content analysis found that it is not a common practice in the industry, with only four identified instances out of the 59 projects where this type of requirement was included in the contract (Table 3.11). Table 3.11 Content Analysis: Warranties Used for Geotechnical Performance Description Freq. Warranties used for geotechnical performance yes 4no 55 In order to seek the early involvement of contractors, owners have the option of including provisions for alternative technical concepts (ATC) in the contract. ATCs allow proposers to

64 submit alternative solutions to potential problems before the contract is awarded. As shown in Table 3.12, the content analysis found that the majority of the projects (32) allowed the submission of ATCs to identify and solve geotechnical issues. Table 3.12 Content Analysis: Provisions for ATCs Description Freq. Provisions for ATCs Yes - can offer different geotechnical design 32 Propriety only 0 Select from approved list 2 No 24 In some projects that are considered to have high geotechnical risks, the owner might include specific methods for mitigating geotechnical conditions in the project as a way to ensure that, if an event occurs, the contractor will solve the problem according to the owner’s requirement. As shown in Table 3.13, the content analysis found projects where this was implemented; however, it is not common practice in DB projects as only eight out of the 59 projects included this type of measure. Table 3.13 Content Analysis: Methods for Mitigating High-risk Geotechnical Conditions Description Freq. Methods for mitigating high-risk geotechnical conditions (e.g., landslides and contaminated soils). yes 8 no 51 The content analysis supports three major findings as well as a number of less important trends.  Despite the lack of a federal mandate requiring a DSC, more than half the documents included one.  The qualifications and past performance of the design-builder’s geotechnical and construction staff were the most often evaluated aspects.  Most of the 59 project documents assigned no weight to geotechnical evaluation criteria.

65 3.3 Summary and Analysis Surveys This section discusses the findings of the two surveys conducted as part of this project. The complete questionnaires are attached in Appendix C. The surveys were developed based on survey methods suggested by Oppenheim (2000). The purpose of the first survey was to identify DOT policies and procedures for articulating geotechnical information and requirements on DB projects. The surveys also furnished real-time perceptional data regarding practitioner definitions of the importance of geotechnical factors associated with geotechnical risk management on typical DB projects. It sought to identify successful approaches for managing geotechnical risks across the DB project’s life cycle as well as discuss those practices that did not adequately address the geotechnical requirements and caused the agency to hold geotechnical liability that it had hoped to shed. Similarly, the purpose of a second questionnaire was to gauge the impact of geotechnical risk factors on DB projects. The obtained results were used to identify those geotechnical factors that could preclude a given project from being delivered using DB. The results are divided into the following categories, for each of the two surveys: 1. Respondent agency demographics. 2. Findings and analysis 3. Conclusions 3.3.1 Findings from Survey of State DOTs (Survey 1) The first survey was addressed to all state DOTs, and was aimed at identifying DOT policies and procedures for articulating geotechnical information and requirements on DB projects.

66 The survey was initially issued to the members of the AASHTO Subcommittees on Construction and Design in each of the 50 DOTs. The subcommittee members were asked to then forward the survey to the person best qualified to respond on an overall department basis. The survey consisted of four major sections; general information about the participant and DB use by the agency, geotechnical risk management information, geotechnical aspects of the DB procurement process, and geotechnical aspects of DB contracts (refer to Appendix F for details on survey design). The findings are organized in the following subsections based on the survey organization. 3.3.1.1 Respondent Agency Demographics Responses were received from 38 DOTs yielding an overall response rate of 76%. Around 80% of the respondents work in the departments’ geotechnical/foundations sections (Figure 3.4). Hence, the survey results reflect the perceptions of the most technically qualified group within each agency. Figure 3.4 Respondent Group/Section Assignment. 2% 10% 79% 3% 3% 3% Design group/section Construction group/section Operations group/section Geotechnical /foundations group/section Alternative project delivery group/section Materials group/section Contracts/procurement group/section Other. Please specify:

67 3.3.1.2 General Information and DOT Alternative Contracting Methods Use and Experience Figure 3.5 shows the contracting methods that the responding 38 states are allowed to use. As seen, all state DOTs are allowed to use DBB. DB Best Value is allowed in 22 states of the 38 states. DB Low Bid is allowed by 17 DOTs, and CMR/CMGC is allowed by 11 DOTs. Ohio and West Virginia DOTs reported the use of Public Private Partnership, while California DOT (Caltrans) reported the use of Design-Sequencing. Connecticut DOT reported that DB and CMGC were only authorized for use in limited number of projects. The state DOTs that did not use DB best value or DB low bid were then allowed to exit the survey with a final hypothetical question regarding their major concern regarding the development of geotechnical requirements for the advertisement and letting of DB contracting, if their DOT would decide to implement DB contracting techniques. The responses, received from the other states centered on the uncertainty of geotechnical conditions prior to bidding, locations and depths of the borings, and whether the test conducted are what the design-builders will need. In addition, concerns were expressed about liability and change orders when actual conditions differ from the owner-supplied information. Figure 3.5 Use of Alternative Contracting Methods 0 5 10 15 20 25 30 35 40  DBB  Construction Manager‐at‐Risk or Construction Manager/General Contractor?  DB Best Value  DB Low Bid  Other No. of DOTs

68 Of the 27 DOTs remaining that use at least a form of DB project delivery, 18 reported delivering more than 10 DB projects, 4 had 6-10 projects, 1 DOT had 3-5 projects, 3 DOTs had 1- 2 projects, and one failed to report this data (Figure 3.6). Figure 3.6 How many DB projects has your agency delivered? As illustrated in Figure 3.7, respondents were asked about the years that the agency has been using DB; 12 DOTs reported using DB for more than 10 years, 10 DOTs between 6-10 years 2 DOTs (Montana and New York) between 3-5 years, and 2 DOTs (Connecticut and Kansas) between 1-2 years. In terms of respondents’ experience with DB projects, 7 of the respondents have more than 10 years of experience, 8 with 6-10 years of experience, 6 with 3-5 years of experience, and 5 with 1-2 years of experience with DB projects. 0 2 4 6 8 10 12 14 16 18 20 1-2 3-5 6-10 >10 No. of DOTs N o. o f p ro je ct s d el iv er ed

69 Figure 3.7 How long has your agency been using DB projects delivery? 3.3.1.3 Geotechnical Risk Management Information This section presents how agencies manage the geotechnical risk in their DB projects. Agencies were asked about the use of a manual or document that specifically describes the procedures to be used with the geotechnical requirements of DB and/or DBB projects. Of 27 DOTs, 11 DOTs had a manual, while 11 DOTs did not use manual, and 33 respondents did not know. Colorado DOT expressed that the manual is in development and is not ready for distribution. Agencies were then asked regarding the amount of preliminary geotechnical investigation performed before making the decision to use DB for a given project. Most of these agencies selected at least two types, indicating that the type of study depends on each project. Out of the 27 DOTs that responded, six DOTs reported not conducting any preliminary investigations (Figure 3.8). Of the remaining DOTs, eight DOTs selected preliminary geotechnical design report. Only Massachusetts DOT reported Geotechnical Interpretation Report. Furthermore, some DOTs like Colorado DOT, for instance, stated that the decision to do geotechnical studies is made by regional program managers, while Maryland SHA pointed out that the use of DB delivery is independent of geotechnical information available. In general, DOTs varied in the amount of geotechnical investigations they performed prior to deciding to use DB. 0 2 4 6 8 10 12 14 1-2 years 3-5 years 6-10 years > 10 years No. of years agency has been using DB N o. o f D O Ts

70 Figure 3.8 How much preliminary geotechnical investigation is completed before making the decision to use DB project delivery for a given project? Of the DOTs that responded that they use DB project delivery on projects where the geotechnical risks are considered to be significant, almost 80% affirmed they used DB and only five DOTs did not. It is notable that DOTs who do not do any preliminary geotechnical studies before making the decision to use DB project delivery, except Hawaii DOT, also stated that they will use DB where the geotechnical risks are considered to be significant. As shown in Figure 3.9, DOTs that use DB project delivery when the geotechnical risk is considered to be significant reported addressing the significant geotechnical issues in their RFP/RFQ mostly by developing GBR (45%), followed by the allowance, restriction, or elimination of DSC rights and mandatory design (20%). 19% of the DOTs reported other measures. For example, Nevada DOT reported allowing short-listed teams to perform their own geotechnical investigations prior to finalizing the RFP and Maine DOT reported allowing for a supplemental boring and lab testing program during the bidding period based on boring requests from bidders/proposers. Kentucky, Maine, Maryland, 0 2 4 6 8 10 None Reconnaissance Report (Review of records and observations from site) Geotechnical Data Report (Review of records and limited investigation data) Geotechnical Summary Report (Review of records and geotechnical investigation of critical areas) Preliminary Geotechnical Design Report (Partial geotechnical investigation) Geotechnical Design Report (Full subsurface investigation for all structures and geotechnical features) Geotechnical Baseline Report (GBR) (A report that establishes the contractual understanding of subsurface… Geotechnical Interpretation Report (GIR) (A report that interprets the findings of the GBR) Other Frequency

71 Minnesota, and Ohio DOTs reported managing these geotechnical issues not only by GBR or by DSC clause, but also by mandatory design. Figure 3.9 Steps Taken to Address Geotechnical Issues in the DB RFQ/RFP Where the Geotechnical Risks Are Considered Significant DOTs who do not use DB project delivery where the geotechnical risk is significant reported reasons such as liability consideration, political issues preventing its use. These responses support the idea that DB delivery selection is mainly governed by schedule and funding instead of geotechnical risk. The survey continued to ask DOTs about geotechnical characteristics or factors that can preclude a given project from being a DB project delivery. The majority of the responses pointed out that geotechnical factors and/or geotechnical risks do not prevent a project from being delivered using DB. New York DOT, for instance, stated that the DB project delivery decision is made by policy makers, not engineers. Regarding geotechnical risk analysis in a typical DB project, respondents were asked about conducting a formal geotechnical risk assessment covering project scope, project schedule, project cost, contracting risk, etc. The responses revealed that most of the DOTs do not conduct a formal risk analysis in any of these areas. The Colorado and Nevada DOTs are the exception, including a formal analysis of geotechnical risk. Also, responses illustrated that a qualitative analysis is more common compared to a quantitative analysis (Figure 3.10). Some comments regarding formal 0 2 4 6 8 10 12 14 16 Geotechnical Baseline Report (GBR) A report that establishes the contractual understanding of… Allowance, restriction or elimination of Differing Site Conditions (DSC) rights Mandatory design Other No. of DOTs

72 geotechnical risk analysis pointed out that risk analysis is only carried out for major projects or those who have a high cost, and that geotechnical aspects are considered within the overall project risk analysis. Figure 3.10 Is a formal geotechnical risk analysis conducted on a typical DB project in any of the following areas? The 14 DOTs, which conduct a formal geotechnical risk analysis were asked to detail their geotechnical risk management process. Figure 3.101 shows that 10 DOTs reported conducting formal risk analysis in typical DB projects, while 7 DOTs reported conducting formal risk analysis in a DB projects with significant geotechnical risk prior to bid. Figure 3.11 also shows that all of the geotechnical risk management processes for DB projects with significant geotechnical risks were conducted either by the agency or required of the design-builder. Only the Georgia and Nevada DOTs responded that they request the design-builder to maintain a risk register during the course of the project that includes anticipated geotechnical risks and mitigation measures. 0 2 4 6 8 10 12 14 16 Q ua lit at iv e Q ua nt ita tiv e N on -F or m al G eo te ch ni ca l R is k A na ly si s Q ua lit at iv e Q ua nt ita tiv e N on -F or m al G eo te ch ni ca l R is k A na ly si s Q ua lit at iv e Q ua nt ita tiv e N on -F or m al G eo te ch ni ca l R is k A na ly si s Q ua lit at iv e Q ua nt ita tiv e N on -F or m al G eo te ch ni ca l R is k A na ly si s Q ua lit at iv e Q ua nt ita tiv e N on -F or m al G eo te ch ni ca l R is k A na ly si s Project Scope Project Schedule Project Cost Contracting Risk Other N o. o f D O Ts

73 Figure 3.11 Within the geotechnical risk management process that is conducted by the agency or required of the design-builder? Regarding risk register content, 8 DOTs reported using both a risk register containing geotechnical risks with probabilistic estimate of cost and schedule impact of risk, and that the risk register developed by the agency determines the risk management mitigation strategies applicable to the geotechnical risks identified (such as share, transfer, and avoid). It should be noted that New York DOT is the only DOT in this pool which doesn’t do preliminary geotechnical investigations before making the decision to use DB delivery nor develop a geotechnical risk register. Additionally, the Colorado, Ohio, nor New Hampshire DOTs also reported no need to develop a risk registers. New Hampshire DOT reported that it does not use DB delivery method on projects where the geotechnical risk is considered to be significant. Figure 3.12 represents the differing risk register content and the number of DOTs using each. 0 2 4 6 8 10 12 Formal risk identification meetings are conducted by the agency's project team prior to bid Risk register, encompassing geotechnical risks, is developed by the agency Risk mitigation report which includes procedures for mitigating risks identified during the risk analysis process is developed Design-builder has to develop a risk management plan to be submitted in the proposal to the agency Design-builder has to maintain a risk register during the course of the project that includes the geotechnical risks anticipated and mitigation measures. No.of DOTs DB Project with Significant Geotechnical Risk DB Typical project

74 Figure 3.12 Which of the following best describes the content of the risk register of geotechnical issues? Figure 3.13 illustrates geotechnical risks encountered during the process and how such risk is allocated. In general, most DOTs allocated the design risks to the design-builder as the contractor is responsible to complete the design. No DOTs took the risk when assumptions are involved; however, incorrect geotechnical design information was the most common risk assumed by the DOTs. Only Hawaii DOT took the risk of ‘bias and or variation in design parameters being different than estimated’. It is noteworthy that DOTs with high seismic risks such us California and Washington, and those who are in the same seismic niche like Missouri, Illinois, Kentucky, and Arizona, did not assumes the risk of seismic design assumptions and recommendations. Only South Carolina, which is considered to have high seismic risk, Kansas, and Massachusetts DOTs allocated this type of risk to the owner. As for the allocation of the geotechnical risk in the procurement phase or in contract, most DOTs either shared the risk via a DSC contract clause or allocated the risk to the design-builders, 0 2 4 6 8 10 Risk register developed by the agency- encompassing geotechnical risks- is maintained during the course of the… Risk register containing geotechnical risks and probabilistic estimate (range) of cost and schedule impact of risk Other Risk register developed by the agency determines the risk management mitigation strategies applicable to the… Risk register developed contains geotechnical risks with a deterministic estimate of the cost and schedule impact of risk No. of DOTs

75 Figure 3.13 What types of geotechnical risks do you typically encounter on DB projects and how are they allocated? Respondents were asked whether their project cost estimates included a quantitative analysis of geotechnical uncertainty. Georgia, Nevada, South Carolina, and Washington DOTs affirmed that they did. Of those, only Nevada DOT does a quantitative formal geotechnical risk analysis in project scope, schedule, cost and contracting risk. Of the four DOTs, Nevada had the least DB experience with the remaining DOTs having completed more than 10 DB projects. As for developing a line item, risk-based, cost estimate of geotechnical risks, only Georgia, Nevada, and South Carolina responded that their quantitative analysis of geotechnical uncertainty includes this factor, while Washington did not. In the last question addressing geotechnical risk management, the respondents were asked if they employed any formalized geotechnical risk allocation technique to draft the contract provision. Five state DOTs (Maine, Massachusetts, Michigan, Minnesota, and South Carolina) out of 18, reported they use a formalized geotechnical risk allocation technique to draft the contract provision. These techniques included sharing the cost sharing and a unit price pay items. 0 5 10 15 20 25 Inadequate geotechnical investigation Incorrect geotechnical design information, in general Bias and/or variation in design parameters being different than estimated, in general Inaccurate earthwork assumptions- soil or rock cuts or fills Risk in retaining structures assumptions and recommendations - geotechnical aspects Risk in structure foundations assumptions and recommendations (footings, driven piles, drilled shafts,… Risk in ground improvement technique recommendations (wick drains, lightweight fill, vibro-compaction, dynamic… Risk in seismic design assumptions and recommendations No. of DOTs OWNER DB CONTRACTOR SHARED

76 Geotechnical Aspects of DB Procurement Process This section of the report presents the results on how DOTs perceive geotechnical factors in DB projects, and details the amount of geotechnical information provided by DOTs. Figure 3.14 shows that 10 DOTs reported assigning a weight to geotechnical factors with regard to all other evaluated factors. The California, Kansas, and New Hampshire DOTs stated that these factors carry no weight. It is notable that only California out of the three DOTs reported using DB delivery on projects where the geotechnical risk is considered to be significant. Virginia DOT was the only DOT to report assigning geotechnical factors a heavy weight compared to other evaluated factors. In general, geotechnical factors in DB RFP evaluation plans are not heavily weighted when compared to all evaluated factors. Figure 3.14 DOTs and Geotechnical Factors Weight in the Evaluation Plan Figure 3.15 shows the geotechnical content that DOTs provide in their RFPs in a DB project with significant geotechnical issues. The majority of the DOTs (Arizona, California, Hawaii, Maine, Michigan, Nevada, New York, North Carolina, and Washington) provide a geotechnical data report. Connecticut, Michigan, Missouri, South Carolina, and Washington DOTs provide geotechnical baseline report in their RFPs, while Arizona, Kansas, and New Hampshire do not provide any geotechnical information in their RFPs. Only Massachusetts DOT answered they provided GIR in their RFPs. Many respondents noted that the type and amount of RFP geotechnical information provided in the RFPs varies; the higher the geotechnical risk, the greater 0 2 4 6 8 10 12 No weight Minor weight Some weight Heavy weight No. of DOTs

77 level of investigation. Other DOTs indicated that all subsurface uncertainties should be addressed in the pre-award phase because after the contract is awarded, the design-builder is responsible to develop geotechnical solutions. Figure 3.15 DOT Survey Response Results Regarding Geotechnical Content Provided in the RFP. Figure 3.16 shows the additional information that design-builders solicit during proposal preparation in a DB project with significant geotechnical issues. As seen, 11 DOTs agree that alternative technical concepts for geotechnical features of work, and proposed mitigation approaches for known or potential risk areas are the most common additional information that design-builders request. Minnesota and Ohio DOTs provide a geotechnical design report as a part of RFPs for projects with significant geotechnical issues and South Carolina provides a GBR. This infers that the more geotechnical information DOTs provide in their RFPs, the less information is required by design-builders. Utah DOT selected “other” option stating “additional subsurface investigation and lab testing required to fulfill (as a minimum) the requirements set forth in our Geotechnical Manual of Instruction; and development of ALL geotechnical design parameters, using both the agencies’ preliminary data report, and the design-builders’ additional investigation.” 0 2 4 6 8 10 None Reconnaissance Report (Review of records… Geotechnical Data Report (Review of records and… Geotechnical Summary Report (Review of records… Preliminary Geotechnical Design Report (Partial… Geotechnical Design Report (Full subsurface… Geotechnical baseline report (GBR) Other No. of DOTs

78 Figure 3.16 RFP Additional Geotechnical Information by Design-Builders. DOTs were asked about what they allow and/or give during the bidding stage for DB projects with significant geotechnical issues. Most DOTs except New York and North Carolina DOTs stated they allow ATCs during the DB procurement (Figure 3.17). However, North Carolina DOT reported that ATCs for geotechnical features or work are required as additional geotechnical information in DB proposals. Around 83% DOTs give the bidders general access to the project site and the results of the borings they had conducted. Figure 3.17 DOTs Procurement Phase Practices in a DB Project with Significant Geotechnical Risks 0 2 4 6 8 10 12 None List of assumptions made regarding geotechnical conditions Limited additional testing as requested by the design- builders Pre-award geotechnical investigation of critical areas by design-builders Geotechnical design values to be used Preliminary designs for foundation features of work Proposed mitigation approaches for known or potential geotechnical risk areas Alternative technical concepts for geotechnical features of work Other No. of DOTs 0 5 10 15 20 25 Allows the proposers to do their own boring at the site Gives the bidders general site access and the results of the borings it had conducted. Allows alternative technical concepts (ATC) during the DB procurement process No. of DOTs

79 Figure 3.18, details the results of the survey with regard to the impact on project success of various aspects that are part of the procurement process. One and see that ‘sufficient geotechnical information to allow the competitors to price the project without excessive contingencies’ was rated the highest with an average of 2.84, followed by ‘highly qualified geotechnical design engineers’ with an average of 2.52. The least important factor was the “peer- review of geotechnical design reports (GDR) and supplemental GDRs” followed by “formal geotechnical risk analysis conducted by agency”, “formalized geotechnical risk allocation techniques to draft the contract provisions”, and “geotechnical risk mitigation plan in proposal.”. Looking closer at the individual DOT responses, only Arizona, Maine, Missouri, Virginia, and Washington rated the “formal geotechnical risk analysis conducted by the agency” as very important. It should be noted that the Maine and Washington DOTs do a qualitative formal geotechnical risk analysis for developing project scope, schedule, cost, and contracting risks. Some outliers were observed in terms of rating the factors’ importance, such as Ohio DOT being the only DOT that selected “geotechnical ATCs with confidential one-on-one meetings” as not important, and Kansas DOT also being the only DOT selecting “correct weight of geotechnical issues in relation to other project requirement area” as not important.

80 Figure 3.18 Average Scores of Importance of Geotechnical Aspects During the Procurement Process to the Overall Success of the Project 3.3.1.4 Geotechnical Aspects of DB Contracts This section reports the results of the last section of the survey that addresses the geotechnical aspects of DB contracts. Respondents were asked to select what type of payment provisions are contained in both typical DB projects and those with significant geotechnical issues. Figure 3.19 shows that lump sum is the most common contract payment provision for both in a project types, with eleven and nine respondent selections respectively, followed by the “combination lump sum and unit price” contract payment provision. The least used types of payment provisions were unit price GMP, unit price, and cost reimbursable, and only the Georgia DOT selecting unit price GMP for DB projects with significant geotechnical issues. 0 1 2 3 Sufficient geotechnical information to allow the competitors to price the project without excessive contingencies Highly qualified geotechnical design engineers Mandated use of agency design criteria Geotechnical ATCs with confidential one-on-one meetings Verification of knowledge and experience working in the project area Opportunity for competitors to conduct some form of subsurface investigation during proposal preparation Correct weight of geotechnical issues in relation to other project requirements Geotechnical design QA plan in proposal Geotechnical construction QA plan in proposal Detailed GBR in RFP Formal geotechnical risk analysis conducted by agency Formalized geotechnical risk allocation techniques to draft the contract provisions Geotechnical risk mitigation plan in proposal Peer-review of GDR and supplemental GDRs No. of DOTs

81 Figure 3.19 Type of Payments in a DB Project DOTs were asked a question on how the contract document is set up to address various geotechnical aspects (with ‘yes’, ‘no’, and ‘don’t know’ response options). A total of 25 DOTs responded to this question. As shown in Figure 3. 20, 20 DOTs (out of 25) stated that they provide and require geotechnical design criteria in their DB contracts, and use performance verification or measurement methods for geotechnical features. This was followed by 19 DOTs who stated that they require geotechnical design criteria in their DB contracts; however, New York and Utah DOTs do not require and provide geotechnical design criteria. The least common way by which DOTs address geotechnical risks in contracts - with more than 50% of the DOTs (14 out of 24 DOTs) responding ‘no’- was having ‘incentives that are used to align owner and contractor geotechnical risks and rewards’. This was followed by the use of warranties which was the second least commonly used. It is interesting to note that South Carolina DOT, for instance, does not use GBR as a contract document, but the agency completes a GBR study before making the decision 0 2 4 6 8 10 12 D B T yp ic al P ro je ct D B P ro je ct w ith S ig ni fic an t G eo te ch ni ca l I ss ue D B T yp ic al P ro je ct D B P ro je ct w ith S ig ni fic an t G eo te ch ni ca l I ss ue D B T yp ic al P ro je ct D B P ro je ct w ith S ig ni fic an t G eo te ch ni ca l I ss ue D B T yp ic al P ro je ct D B P ro je ct w ith S ig ni fic an t G eo te ch ni ca l I ss ue D B T yp ic al P ro je ct D B P ro je ct w ith S ig ni fic an t G eo te ch ni ca l I ss ue D B T yp ic al P ro je ct D B P ro je ct w ith S ig ni fic an t G eo te ch ni ca l I ss ue D B T yp ic al P ro je ct D B P ro je ct w ith S ig ni fic an t G eo te ch ni ca l I ss ue Lump sum Lump sum guaranteed maximum price (GMP) Unit price GMP Unit price Cost reimbursable Combination lump sum and unit prices Other N o. o f D O Ts

82 to use a DB project and also provides a GBR study in the RFP in a project with significant geotechnical issue. On the other hand, Michigan, Missouri, and Washington DOTs, for example, are agencies that not only use GBR as a contract document, but also include the GBR in their RFP for a DB project with a significant geotechnical issue. Figure 3. 20 Geotechnical Aspects of DB Contracts When asked whether their DB contracts contain a clause regarding geotechnical differing site conditions, 89% reported that they did. The Arizona, Kansas, and Ohio DOTs reported they did not. Of the three, only the Ohio DOT uses DB delivery on projects with significant geotechnical and on those, the agency addresses the geotechnical issue in RFP/RFQ by including a GBR and prescriptive design criteria. The agencies that used a geotechnical differing site conditions clause were asked to rate how often design-builders’ claims of a differing geotechnical site condition result in a compensable change order and more than 50% of the DOTs responded “occasionally” while none of these marked “always”. Connecticut, Louisiana, Maine, Maryland, New York, and Utah DOTs responded that they have “never” had a claim that would result in a compensable

83 change order. Of that group, the Connecticut DOT is the only agency that does not use DB delivery when the geotechnical risks are considered to be significant. In contrast, Hawaii, Kentucky, and Massachusetts DOTs selected “usually” for a frequency of design-builder’s DSC claims. It should be noted that Massachusetts DOT is the only agency that allocates the risk of differing site conditions contract clause to the owner. Respondents whose contracts contain a clause regarding geotechnical differing site condition were further asked if the clause explicitly delineate the contractor’s right to submit a claim for specific types of unforeseen conditions. California, Connecticut, Maine, Michigan, Minnesota, Nevada, South Carolina, Virginia, and Washington DOTs affirmed the clause did, while Massachusetts and New Hampshire DOTs reported the clause did not, and Georgia, Hawaii, Kentucky, Louisiana, Maryland, Missouri, Montana, New York, North Carolina, and Utah DOTs reported “no opinion”. Some DOTs clarified that differing site conditions expressly do not include conditions related to geology or hydrology; it includes bedrock, soils, groundwater, or other natural conditions. It was also noted by one DOT that contracts typically state that geotechnical conditions are not considered a differing site conditions. Next, DOTs that work with contracts containing a differing site conditions clause were asked to estimate the percentages of their DB projects that have ended up with a compensable differing site conditions change/claim. The percentages were divided into five categories, none; 1- 10%; 11-25%; 26-50%, and more than 50%. Only California DOT selected more than 50%. It is interesting that Caltrans was the only agency that reported that more than half of its DB projects have ended up with a compensable differing site condition with the agency having more than 10 DB delivered and more than 10 years of having using DB as a delivery method. This answer can be explained based on the fact that Caltrans neither allocates any weight to geotechnical factors in

84 the evaluation plan, nor completes a formal geotechnical risk analysis. On the other hand, Connecticut, Louisiana, Maine, Maryland, New York, and Utah marked none as a percent of its projects. Kentucky, Massachusetts, and Nevada DOTs marked 11-25%. From 19 DOTs that responded to this question, none selected 26-50%, and nine DOTs marked 1-10%. Agencies were asked to select what document, if any, is used to define a differing geotechnical site conditions. There were six options where “no document” and “other options” were included as an answer (Figure 3.21). Most DOTs (7 DOTs) reported that the contract differing site condition clause was the only source. Four DOTs (California, Maryland, North Carolina, and Utah DOTs) selected the geotechnical information contained in the RFP, while the same number of DOTs (Connecticut, Missouri, Virginia, and Washington) selected GBR contained in the RFP.. Figure 3.21 What document, if any, is used to define a differing geotechnical site condition? To seek potential case studies for DB projects involving geotechnical issues, the questionnaire asked DOTs if their agency had a major claim regarding geotechnical problems. Seven DOTs responded that they did; those included California, Georgia, Maine, New Hampshire, Utah, Virginia, and Washington. That group indicated that geotechnical issues included a failure of foundation due to inaccurate subsurface information, unexpected obstruction encountered 0 1 2 3 4 5 6 7 8 Geotechnical information contained in RFP GBR contained in RFP GDR produced by design-builder Contract differing site conditions clause definition only No document Other

85 during mini-pile installation, karst terrain, landslide during construction, and tunnel launch pit dewatering. In order to compare DB and DBB projects, agencies were asked to rate the final quality of geotechnical work on DB projects compared to DBB projects. The options were: Better, Same, Worse, and No Opinion (Figure 3.22). From 25 DOTs, only South Carolina marked “Better” providing justifications such as “our DB projects are generally higher profile and higher complexity projects. Because a quality component is included in both the short-listing of the DB teams in the RFQ process and quite often in price adjustments based on a quality score in the RFP and bid process, there is an incentive for the contractors to assemble the best quality team possible”. In addition, it is notable that South Carolina is one of the three DOTs that perform a complete GBR investigation before making the decision to use DB as a delivery method, it is also one of the three DOTs that bear the risk of inadequate geotechnical investigation, and is one of the five DOTs that provides GBR in a project with significant geotechnical issue. Thirteen DOTs rated the quality of geotechnical work to be the same as DBB projects, while California, Kentucky, Montana, New York, Ohio, Utah, and Washington DOTs considered it “worse”. Some reasons provided as to why the final work of geotechnical quality is worse were as follows:  Contractor-designers like consulting engineers are unwilling to take on any risk at all in their deigns, thus the agency gets functional but conservative and expensive geotechnical solutions.  Contractors want to “cheapen up” the geotechnical work; and sacrifice of some degree of quality due to the lack of authority of the agency over design-builder.  Colorado, Connecticut, Maryland, and Nevada selected “no opinion.”

86 It should be noted that South Carolina DOT formally assesses design-builder’s performance and uses that for the future DB selection. SCDOT was the only agency that selected better on comparing the final quality of geotechnical work on DB projects with DBB projects. Figure 3.22 How do you rate the final quality of geotechnical work on DB projects compared to DBB projects? The next question asked if agencies formally evaluated the design-builder’s performance quality and use that for future DB selections. From 25 DOTs that responded to this question, only Connecticut and South Carolina marked “yes.” The remaining responses were almost equally divided between “no” and “do not know” answers; twelve and eleven DOTs marked “no” and “do not know”, respectively. The last question regarding geotechnical aspects of DB contracts asked agencies to rate 11 geotechnical factors from 1 to 5 for their impact on the final quality/performance of the project, with 1 being very low impact to 5 being very high impact. Figure 3.23 shows that the highest factors rated were the “qualifications of the design-builder’s geotechnical staff” and the “agency interactivity with geotechnical design team during design phase”. This was followed by the “amount of geotechnical information expressed in procurement documents” and “design-builders past project experience with geotechnical issues”. The lowest factor rated was “warranty provisions.” 0 2 4 6 8 10 12 14 Better Same Worse No Opinion

87 Figure 3.23 Please rate the following geotechnical factors for their impact on the final quality/performance of the DB project. Finally, the questionnaire asked DOTs that work with DB project delivery method if there is something that they would like to share regarding geotechnical aspects on their DB projects. One of the main recurring comments was the importance of having adequate exploration that defines the likely geotechnical issues and allows the DB contractors to properly scope and bid the project. Comparing Experienced and Non-Experienced DOT’s Geotechnical Risk Management Strategies The research team decided to take a closer look at the results detailed above, pertinent to the experience of the different DOTs in DB project delivery and categorized the respondent into experienced and non-experienced (Table 3.14 - Survey Respondent and Categorization Based on Experience in DB Projects), to investigate the differences between these two groups in managing 1 1.5 2 2.5 3 3.5 4 4.5 5 Qualifications of the Design-Builder's geotechnical staff Agency interactivity with geotechnical design team during design phase Amount of geotechnical information expressed in the procurement documents Design-Builder's past project experience with geotechnical issues Use of geotechnical performance criteria/specifications Use of agency geotechnical specifications and/or design details Early contractor involvement in geotechnical design Geotechnical ATCs Agency interactivity with geotechnical design team during proposal phase Confidential one-on-one meetings Warranty provisions Average Score with 5 being

88 geotechnical risks on their DB projects. Experienced agencies are defined as those having completed more than 10 DB projects. Table 3.14 Survey Respondent and Categorization Based on Experience in DB Projects Category Responding States Experienced DOTs ( >10 DB projects) Arizona, California, Colorado, Georgia, Kentucky, Maine, Maryland, Michigan, Minnesota, Missouri, Montana, New York, North Carolina, Ohio, South Carolina, Utah, Virginia, Washington, West Virginia Non-experienced DOTs ( <10 DB projects) Connecticut, District of Columbia, Hawaii, Kansas, Louisiana, Massachusetts, Nevada, New Hampshire This investigation led to some major findings, as detailed below: Many DOTs believe that contractors should assume full risk of differing site conditions (Christensen and Meeker 2002). However, the results of the online-survey for this research show that not all agencies share the same thought. Particularly, the study found that experienced and non-experienced DOTs allocate unknown geotechnical conditions differently. Experienced agencies tend to share geotechnical risk; whereas non-experienced agencies tend to either accept or shed all of it. In terms of the weight of geotechnical factors, it was observed that while more than 80% of non-experienced DOTs indicated that geotechnical factors have “No/Minor Weight” in their evaluation plans. However, more than 60% of the experienced DOT’s placed a Some/Heavy weight to geotechnical factors (Figure 3.24). This leads one to infer that experienced DOTs realize the importance of giving weight to geotechnical factors in DB evaluation plans.

89 Figure 3.24 Geotechnical Evaluation Criteria Weighting As for the three geotechnical factors, shown in Figure 3.25, in terms of their impact on the final quality/performance of the DB project, a consistent difference is observed in the higher consideration of experienced DOTs when compared with the non-experienced DOTs. For instance, 100% of experienced DOTs not only agree in rating the qualification of the design-builder’s geotechnical staff as a very/high impact, but also agree in rating the design-builder’s past project experience with geotechnical issues as very/high impact. In contrast, non-experienced DOTs results are dispersed across the range of impacts. Non-experienced DOTs do not perceive the importance of qualifying the design-builder’s geotechnical experience and workforce as being as important as the experienced DOTs. Experienced DOTs see the amount of geotechnical information included in RFPs as important thereby decreasing geotechnical uncertainty during procurement, and making the proposals received more competitive (Christensen and Meeker 2002). No / Minor Weight Some / Heavy Weight Less than 10 DB projects delivered 83.3% 16.7% More than 10 DB projects delivered 37.5% 62.5% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% Pe rc en ta ge o f A ge nc ie s

90 Figure 3.25 Impact of Geotechnical Risk Factors Further statistical analysis was conducted using the Pearson Chi-Square Test to see if there is a statistical significant difference in the perception of DB’s geotechnical aspects between the two groups. Table 3.15 shows the different aspects tested. A significant difference was observed in the importance of the qualifications and past experience of the design-builder’s geotechnical staff. Table 3.15 Summary of Chi-Square Test Results for Experienced vs Non-experienced DOTs Aspects tested Chi-Square ( ) Significance (p) Impact of the design-builder's geotechnical staff qualifications on DB project quality/performance. 8.327 0.01* Impact of the design-builder's past project experience with geotechnical issues on DB project quality/performance. 11.657 0.00* Impact of the amount of geotechnical information expressed in the procurement documents on DB project quality/performance 2.906 0.08 Geotechnical factor weighting in the evaluation plan compared with other evaluation factors 3.667 0.05 Unknown geotechnical condition risk allocation 4.200 0.12 Another interesting difference between experienced and non-experienced DOTs is related to the allocation of geotechnical uncertainty. Figure 3.26 illustrates that while 57% of experienced Less than 10 DB projects delivered More than 10 DB projects delivered Less than 10 DB projects delivered More than 10 DB projects delivered Less than 10 DB projects delivered More than 10 DB projects delivered Qualifications of the Design-Builder's geotechnical staff Design-Builder's past project experience with geotechnical issues Amount of geotechnical information expressed in the procurement documents Very High/ High Impact 57.1% 100.0% 42.9% 100.0% 57.1% 88.2% Some Impact 28.6% 0.0% 28.6% 0.0% 42.9% 11.8% Slight/No Impact 14.3% 0.0% 28.6% 0.0% 0.0% 0.0% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% Pe rc en ta ge o f A ge nc ie s

91 DOTs are more willing to share the geotechnical risk uncertainty rather than bear it, more than 50% of non-experienced DOTs are willing to bear the geotechnical risk rather than share it, and only 14% of these DOTs allocate it to the owner. This could be attributed to the better understanding of the DB project delivery method as compared to the traditional methods in terms of risk allocation. Figure 3.26 Geotechnical Risk Allocation Finally, Figure 3.27 shows another prominent difference between experienced and non- experienced agencies, which is the perception of the level of importance of completing a formal geotechnical risk. Over 70% of experienced DOTs rated a formal geotechnical risk analysis as either “important” or “very important,” while 50% of non-experienced DOTs gave the formal analysis of geotechnical risk “no importance.” Owner DBContractor Shared Unknown geotechnical condition Less than 10 DB projects delivered 57.1% 14.3% 28.6% More than 10 DB projects delivered 14.3% 28.6% 57.1% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% Pe rc en ta ge o f A ge nc ie s

92 Figure 3.27 Formal Geotechnical Risk Analysis 3.3.1.5 Consolidated DOT Survey Findings Based on the survey findings and the comparison between experienced and non-experienced DOTs, the following conclusions are drawn: Before Making a Decision to Use DB Project Delivery 1. Geotechnical factors, as well as geotechnical information available before making the decision to use DB delivery, not only are independent of the use of DB delivery in agencies, but also do not preclude a given project from using a DB contract. This is supported by the fact that most agencies reported using DB project delivery even when geotechnical issues are considered significant. It was, however, noted that agency direction, policy, schedule, and funding are factors that could preclude a project from being DB. 2. In terms of geotechnical investigation prior to bidding, 76% of DOTs reported conducting a preliminary geotechnical investigation before making the decision to use the DB delivery method. It is interesting that 7 DOTs reported not performing any geotechnical investigations prior to making the decision to use DB, and 80% of the DOT respondents reported using DB in projects with significant geotechnical risks. Very Important Important Not Important/Does not apply Formal geotechnical risk analysis conducted by agency Less than 10 DB projects delivered 0.0% 50.0% 50.0% More than 10 DB projects delivered 29.4% 41.2% 29.4% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% Pe rc en ta ge o f A ge nc ie s

93 3. A formal geotechnical risk analysis affords the DOT project staff the ability to identify, quantify and mitigate the geotechnical risk before the procurement process starts and thus adds value to the DB project delivery process. 4. US case law shows that the owner cannot fully shed geotechnical risk by relying on exculpatory DSC clauses alone. Therefore, geotechnical risk must be confronted early in the project development process, producing thoughtful geotechnical risk management plans that rationally share the project-specific risks with the competing design-builders during procurement and after award of the DB contract. Included in DB Solicitation Documents 5. Uncertainty is reduced by increasing the amount of information available at the time a decision is made. Therefore, the DOT should provide as much information on subsurface conditions as it has at the time the DB project is advertised to permit competing DB teams the greatest information benefit as they make risk pricing decisions during bidding. 6. Majority of the DOTs provide a geotechnical data report followed by preliminary geotechnical design reports in their RFPs. 7. States that reported using DB in projects with significant geotechnical risks mostly used Geotechnical Baseline Reports (GBR) to address geotechnical issues in the RFQ/RFP. 8. Results show that geotechnical factors, if included in the evaluation plan, carry at least some weight with regard to all other evaluated factors. However, agencies are aware that geotechnical factors have an impact not only on schedule and cost of the project, but also on project’s overall success.

94 9. One major concern that was repeatedly reported by DOTs related to the success of the project is to have sufficient geotechnical information during the procurement phase to allow competitors to price the project without excessive contingencies. 10. Based on the comparison conducted between experienced and non-experienced DOTs in managing geotechnical risks, it was seen that more experienced DOTs give more emphasis to the qualifications and past experience of the design-builder’s geotechnical staff and their effect on the final quality/performance of the DB project. 11. The cost of geotechnical uncertainty at the time of bidding can be mitigated by the thoughtful allocation and sharing of project-specific geotechnical risks detailed in the DB RFP. During Bidding 12. Supplemental studies such as allowing proposers to perform their own geotechnical investigations and laboratory testing, prior to the bid, are steps that agencies implement to address geotechnical issues in the DB RFQ/RFP. 13. There is a general agreement in regards to the importance of the interactivity between the agency and the design-builder prior to the bid. This is evident by many agencies requesting ATCs, which propose mitigation approaches for known or potential geotechnical risk areas. 14. In general, the geotechnical risk management process conducted by the agency, prior to bid, is more common in DOTs rather than a risk management process required of the design-builder.

95 15. Based on the results, it seems that DOTs, in general, do not conduct a formal geotechnical risk analysis in the preparation of the project scope, schedule, and costs. When a geotechnical risk analysis is conducted, a qualitative analysis predominates over quantitative analysis. 16. In general, DOT project cost estimates do not involve a quantitative analysis of geotechnical uncertainty supporting that qualitative analysis of geotechnical risk is more likely developed. 17. Unknown geological conditions were reported as the most typical risk encountered on DB projects, with it mostly shared between agency and design-builder. Settlement was the next highly reported form of risk in ground conditions and was mostly borne by the design- builder. In terms of design uncertainty, incorrect geotechnical information was reported as the highest risk, with it being mostly borne by the design-builder. In terms of risk allocation in the DSC clause, most DOTs had the clause drafted to share or transfer DSC risk to the design-builder. 18. Agencies generally are not willing to bear geotechnical risk during the post-award design process intending to make the design-builder responsible for developing geotechnical solutions after the contract is awarded. 19. In general, DOTs are not only providing, but also requiring geotechnical design criteria in their DB contracts. Project Execution 20. Less than 10% of the DOT DB projects end with a claim regarding differing site condition. 21. Only 4 DOTs formally evaluate the design-builder’s performance and use this information for future DB selection. However, most DOTs agree that highly qualified geotechnical Geotechnical Risk Management Process and Risk Allocation

96 design engineers are important to the success of the project. Also that the qualifications of the design-builder’s geotechnical staff and its past project experience with geotechnical issues have a high impact on the final quality/ performance of the DB project. 3.3.2 Survey of DB Geotechnical Risk Experts (Survey 2) The second survey was conducted to gain an expert perception on the frequency and impact of geotechnical risk factors on DB projects. It was conducted as a follow-up on the results of Survey 1 which only included 11 common elements of geotechnical uncertainty typically found on DB projects. This survey was also combined with data from the content analysis of DOT geotechnical design manuals that gained from subject matter experts interviewed in conjunction with the development of agency case studies. The result was an expanded list of 27 final geotechnical risk factors. Two examples of the documents reviewed for the content analysis are Checklist and Guidelines for Review of Geotechnical Reports and Preliminary Plans and Specifications (FHWA 2003) and the GeoConstructability Report (GeoConstructability 2011). These were used to identify geotechnical factors that were not included in first 11-factor list from Survey 1. The survey asked respondents to rate the frequency of 27 risk factors on DB transportation projects using a Likert scale for occurrence and impact ranging from never occurs (0) to occurs very often (4) and from no impact (0) to catastrophic impact (4). Because of the highly specific information sought in the second survey, a pilot questionnaire was sent to one DOT and one industry expert to test the understandability and clarity of the survey. The detailed questionnaire is shown in Appendix F. The pilot test resulted in important modifications to the questionnaire. The survey measured the perceptional differences using DB geotechnical factors impact and

97 frequency assessment and to provide the necessary data to develop an objective ranking of those factors based on importance index theory (Assaf and Al Hejji 2006). 3.3.2.1 Respondent Demographics This survey was sent to DOT and industry geotechnical professionals with DB experience. A total of 46 valid responses were received from 22 DOTs and 24 industry experts, yielding an overall response rate approximately 31%. 3.3.2.2 Results and Analysis of the Expert Survey The analysis of the data was conducted in accordance with importance index methodology (Assaf and Al Hejji 2006; Santoso and Soeng 2016). This approach requires the analyst to initially calculate a frequency index and an impact index based on the perceptions of the expert survey respondents. These are then combined mathematically to calculate an importance index for each factor. The underlying theory is that a risk with a high impact that frequently occurs is more important than a low impact risk that rarely occurs. These indices were then used to objectively rank geotechnical risk factors in DB projects. 3.3.2.3 Frequency of Geotechnical Risk Factors Table 3.16 shows the results of the experts’ perception of the frequency of the 27 factors from the survey. Results were compared against the output from the initial survey which contained 11 of the 27 factors. From both owner and industry perspectives, groundwater/water table issues were rated as the most frequent geotechnical factor. It is notable that the three geotechnical factors (groundwater/water table, settlement, and contaminated material) from the initial survey are within the top 11 factors shown in Table 3.16. In addition, it should be noted that the Table 3.16 factors range from the 20% to 68% frequency occurrence, meaning that subsurface conditions and

98 underground construction activities risks are broadly recognized as being one of the major areas of risk exposure in DB contracts to all parties (Dwyre 2012). Table 3.16 Frequency of Geotechnical Factor Risk in DB Projects FACTOR/RISK Frequency Index [%] Freq. Rank 11-Factor List Rank Water table changes 67.56 1 2 Settlement in general 60.93 2 1 Settlement of bridge approaches 56.82 3 NA Soft compressible soil 55.35 4 NA Replace in situ material with borrowed material 54.76 5 NA Presence of rock/boulders 54.42 6 NA Scour of bridge piers 54.00 7 NA Unsuitable material 53.02 8 NA Contaminated material 51.43 9 4 Ground water infiltration 51.16 10 NA Seismic Risk 50.73 11 NA Highly compressive soils 49.33 12 NA Existing structures likely to be impacted by the work (other than utilities) 48.84 13 NA Slope instability 48.37 14 3 Soft clays, organic silts, or peat 48.26 15 2 Underground manmade debris 47.73 16 NA Settlement of adjacent structure 46.36 17 NA Sensitiveness of public consideration (Parks, historic building, etc.) 46.19 18 NA Liquefaction 44.29 19 NA Landslides 43.00 20 6 Lateral spreading 42.93 21 NA Rock faults/ fragmentation 42.93 22 7 Eroding/mobile ground conditions 39.51 23 NA Subsidence (subsurface voids) 39.07 24 5 Karst formations 38.60 25 9 Caverns/voids 37.62 26 NA Chemically reactive ground 34.87 27 8 3.3.2.4 Impact Geotechnical Risk Factors Table 3.17 shows the impact of geotechnical factors in DB project cost and schedule performance. It is striking that the top five factors are difficult to quantify prior to the execution of

99 the project. These factors, in order, are contaminated material, slope instability, landslides, settlement of adjacent structure, and highly compressive soils. Table 3.17 Impact of Geotechnical Factor Risk in DB Projects FACTOR/RISK Impact Index [%] Impact Rank 11-Factor List Rank Contaminated material 72.38 1 4 Slope instability 72.09 2 3 Landslides 71.50 3 6 Settlement of adjacent structure 68.64 4 NA Highly compressive soils 68.44 5 NA Subsidence (subsurface voids) 66.98 6 5 Soft clays, organic silts, or peat 66.52 7 2 Sensitiveness of public consideration (Parks, historic building, etc.) 65.71 8 NA Scour of bridge piers 65.50 9 NA Soft compressible soil 65.12 10 NA Seismic Risk 63.90 11 NA Existing structures likely to be impacted by the work (other than utilities) 63.72 12 NA Water table changes 63.11 13 2 Lateral spreading 62.93 14 NA Liquefaction 62.86 15 NA Caverns/voids 62.38 16 NA Karst formations 62.33 17 9 Settlement in general 61.86 18 1 Settlement of bridge approaches 61.36 19 NA Rock faults/ fragmentation 60.98 20 7 Underground manmade debris 60.91 21 NA Presence of rock/boulders 60.00 22 NA Unsuitable material 58.60 23 NA Chemically reactive ground 57.95 24 8 Eroding/mobile ground conditions 56.59 25 NA Ground water infiltration 56.28 26 NA Replace in situ material with borrowed material 51.43 27 NA Generally, these factors are encountered after the design-builder has commenced construction on site. One can see that the results of the impact index calculation are generally higher than the frequency index calculations. The impact indices range from the 51% up to 72%.

100 The results are consistent with the findings of NCHRP Synthesis 429 (Gransberg and Loulakis 2012), affirming that geotechnical conditions not only have an enormous impact on DB projects, but also directly affect project cost and schedule. 3.3.2.5 Ranking the Importance of Geotechnical Factors The importance index of each geotechnical risk factor is shown in Table 3.18 in descending order. It is noteworthy that the top two geotechnical factors importance index which are groundwater/water table and settlement in general, are also the first two in the frequency index ranking suggesting that the majority of DB projects are exposed to these two types of risks. It can then be inferred that certain geotechnical factors should be addressed and clarified during the preliminary geotechnical investigation of a DB project. Furthermore, it is interesting note that the first factor (groundwater/water table) based on importance index, seems to be an outlier in regards to all factors. This corroborates the notion that groundwater is an inherent geotechnical factor in transportation projects.

101 Table 3.18 Importance Index of Geotechnical Factor Risk in DB Projects FACTOR/RISK Importance Index [%] Rank Groundwater/ Water table 42.64 1 Settlement in general 37.69 2 Contaminated material 37.22 3 Soft compressible soil 36.04 4 Scour of bridge piers 35.37 5 Slope instability 34.87 6 Settlement of bridge approaches 34.87 7 Highly compressive soils 33.77 8 Presence of rock/boulders 32.65 9 Seismic risk 32.42 10 Soft clays, organic silts, or peat 32.10 11 Settlement of adjacent structure 31.82 12 Existing structures likely to be impacted by the work (other than utilities) 31.12 13 Unsuitable material 31.07 14 Landslides 30.75 15 Sensitiveness of public consideration (Parks, historic building, etc.) 30.35 16 Underground manmade debris 29.07 17 Ground water infiltration 28.79 18 Replace in situ material with borrowed material 28.16 19 Liquefaction 27.84 20 Lateral spreading 27.01 21 Rock faults/ fragmentation 26.17 22 Subsidence (subsurface voids) 26.17 23 Karst formations 24.06 24 Caverns/voids 23.47 25 Eroding/mobile ground conditions 22.36 26 Chemically reactive ground 20.21 27 3.3.2.6 Differences Between Agency and Industry Perceptions of the Impact of Geotechnical Risk Factors in DB Projects In addition to ranking risks, it was also important to identify and report the differences between agency and industry perceptions of the impact of geotechnical risk factors in DB projects. Figure 3.28 shows the results based on the impact index between DB industry and DOTs, and Table 3.19 presents each factor by its corresponding number. Figure 3.28 shows the similarities and differences with respect to geotechnical impact. It is interesting to note that the DB industry’s perspective of overall geotechnical risk factors results in medium impacts with a

102 trend towards high impact, while DOTs’ view the majority of these factors to have low impact. Only “replace in situ material with borrowed material” and “unsuitable material” are common factors belonging to the low impact zone, while “contaminated material” and “landslides” are the highest impact from only DB industry’s perspective. “Slope instability” seems to be a concerning factor for both parties. It is striking that caverns/voids have opposing views from the DB industry and DOTs, with the DB industry rating it a medium/high impact, while DOTs rate it as the lowest impact factor. Figure 3.28 Perception of Geotechnical Risk Impact 2.00 2.50 3.00 3.50 4.00 4.50 2.00 2.50 3.00 3.50 4.00 4.50 IN D U ST R Y AGENCIES H IG H IM PA CT M ED IU M I M PA CT LO W I M PA CT LOW IMPACT MEDIUM IMPACT HIGH IMPACT 1 2 9 8 7 6 5 4 3 13 12 1110 18 17 16 15 14 23 22 21 20 19 27 26 2524

103 Table 3.19 List of Geotechnical Risk Factors Factor/Risk Number Factor/Risk Number Caverns/voids 1 Subsidence (subsurface voids) 15 Chemically reactive ground 2 Existing structures likely to be impacted by the work (other than utilities) 16 Liquefaction 3 Contaminated material 17 Karst formations 4 Landslides 18 Rock Faults/ Fragmentation 5 Settlement of adjacent structure 19 Lateral spreading 6 Sensitiveness of public consideration (Parks, historic building, etc.) 20 Seismic Risk 7 Soft compressible soil 21 Underground manmade debris 8 Groundwater/ Water table 22 Ground water infiltration 9 Settlement in general 23 Presence of rock/boulders 10 Soft clays, organic silts, or peat 24 Settlement of bridge approaches 11 Highly compressive soils 25 Eroding/mobile ground conditions 12 Scour of bridge piers 26 Replace in situ material with borrowed material 13 Slope instability 27 Unsuitable material 14 Considering that DOTs and DB industries are both exposed to some degree of DB geotechnical risk, the difference in the frequency results, were fairly similar between the two parties. Seismic risk, liquefaction, and lateral spreading are the highest and had the widest difference in perceived frequency. This difference is probably because most of the DOT respondents were located in regions of the country where the likelihood of an earthquake is low. Whereas, most of the industry experts work in multi-state areas, which logically increases their exposure to gauging seismic risk on their DB projects. The computation of the importance index is shown in Figure 3.29. One can see that perceptions differ between the DB industry and DOTs. Industry perceives that most geotechnical risk factors are more important than the DOTs. Although the DB industry and DOTs differ on level of importance, both groups agree that the top risk factors are “groundwater/water table,” “settlement in general,” and “scour of bridge piers.”

104 Figure 3.29 Importance Index of Geotechnical Risk Factors Similarly, “eroding/mobile ground condition” and “chemically reactive ground” are factors that have the lowest importance for both groups. The DOTs and industry only rated “replace in IMPORTANCE INDEX [%] 47.00-51.00 Groundwater/ Water table Seismic Risk Settlement in general Contaminated material Scour of bridge piers Soft compressible soil Settlement of bridge approaches Settlement of adjacent structure Liquefaction Ground water infiltration Presence of rock/boulders Underground manmade debris Soft clays, organic silts, or peat Sensitiveness of public consideration Groundwater/ Water table Slope instability Landslides Lateral spreading Highly compressive soils Existing structures likely to be impacted by the work (other than utilities) Settlement in general Rock Faults/ Fragmentation Soft compressible soil Subsidence (subsurface voids) Slope instability Unsuitable material Settlement of bridge approaches Karst formations Scour of bridge piers Caverns/voids Highly compressive soils Presence of rock/boulders Unsuitable material Sensitiveness of public consideration Replace in situ material with borrowed material Soft clays, organic silts, or peat Existing structures likely to be impacted by the work (other than utilities) Contaminated material Landslides Underground manmade debris Eroding/mobile ground conditions Settlement of adjacent structure Chemically reactive ground Ground water infiltration Subsidence (subsurface voids) Rock Faults/ Fragmentation Eroding/mobile ground conditions Karst formations Seismic Risk Liquefaction Lateral spreading Chemically reactive ground Caverns/voids INDUSTRY PERSPECTIVESIMILAR PERSPECTIVEDOTs PERSPECTIVE 15.00-19.00 23.00-27.00 35.00-39.00 27.00-31.00 39.00-43.00 43.00-47.00 31.00-35.00 19.00-23.00

105 situ material with borrowed material” at the same level of importance. On the other hand, Figure 3.29 shows that “seismic risk” had the highest difference of perceived risk. As previously mentioned the difference may be due to the effect of the frequency index calculated for those agencies located in regions with low earthquake risk. A statistical analysis was further conducted using the pooled t-Test to determine if there is a statistically significant difference between the perception of DOTs and DB industry in evaluating the relative magnitude of the 27 geotechnical risk factors in DB projects. Because of the bias found in perceptions of the “seismic” risk factor due to locational impacts, it was dropped from the analysis to give more rigor to the study and provide a more generalizable result. Table 9 shows the two aspects tested. The hypotheses tested were the following: Ho1: There is no statistical difference between DOTs and DB industry in their perception of geotechnical risk factors in DB projects. Ha1: There is a statistical difference between DOTs and DB industry in their perception of geotechnical risk factors in DB projects. The null hypothesis was rejected given that p-value < 0.05 (pooled t-Test = 6.616 and p- value = 0.0001). Thus, there is statistically significant difference between how DOTs and the DB industry perceive geotechnical risk factors in DB projects, with the DB industry perceiving more importance on geotechnical factors. Another purpose of the DB geotechnical factors assessment survey was to identify those factors identified as having “major” or “catastrophic” impact on DB projects, which may cause a DOT expert to recommend foregoing DB delivery and/or an industry practitioner to choose not to bid on a DB project. As seen from Survey 1, geotechnical risk is not currently a critical factor in a DOT’s DB project delivery selection decision. In other words, DOTs will use DB delivery

106 regardless of the level of geotechnical risk. However, the importance index analysis results lead to recommending that the prioritized list of geotechnical risk factors can be used to guide the preliminary investigations for inclusion in the DB RFP. In this context, knowing the geotechnical risk factors that are perceived as critical by the DB industry provides the DOT an understanding of whether to include specific investigations to increase the amount of subsurface information in those areas. This constitutes the pre-award mitigation of the procurement risk of excessive contingencies, which may prevent a DB project award because the bids are too high (Gransberg and Loulakis 2012). Table 3.20 presents a comparison of perceived DOT and industry geotechnical factor risks, which leads one to conclude that the perceptions of critical geotechnical risks is not aligned. Thus, there is a high potential that DOT contingencies will be inadequate as a result of differing perceptions. Christenson and Meeker (2002) would also argue that the pool of potential competitors will be shallower due to the misalignment of perceived geotechnical risk. Table 3.20 Geotechnical Risk Factors that Would Make Not to Pursue/Recommend a DB Project Geotechnical Risk Factor DOTs and DB Industry Agreement DOTs and DB Industry Disagreement Landslides X Subsidence (subsurface voids) X Contaminated material X Prediction of subsurface condition due to inaccessible drilling locations X Sensitiveness of public consideration (Parks, historic building, etc.) X Karst formation X Slope instability X 3.3.2.7 Geotechnical Risk Sharing Mitigation Identification The initial survey asked DOTs respondents how geotechnical risks are typically allocated in DB project RFPs. The ten most common geotechnical risks encountered in DB projects are

107 shown in Table 3.21. The ten factors were then plotted against the differences in the importance index of the two groups found in Figure 3.30, which illustrates the percentage of risk allocation versus the difference in importance index for the ten factors. It also indicates that DOTs shed nine of the 10 most important geotechnical risks. Only “contaminated material” is shared by the DOTs. Settlement in general is the highest geotechnical risk factor that is allocated to the DB builder. Table 3.21 Ten Most Encountered Geotechnical Risk Factors Factor/Risk Number Factor/Risk Number Slope instability 1 Landslides 6 Soft clays, organic silts, or peat 2 Rock Faults/ Fragmentation 7 Chemically reactive ground 3 Settlement in general 8 Subsidence (subsurface voids) 4 Contaminated material 9 Groundwater/Water table 5 Karst formations 10 Figure 3.30 Geotechnical Risk Allocation in DB Contract. Figure 3.30 graphically illustrates an opportunity to mitigate geotechnical risk prior the award of a DB contract. A high difference in importance index means that the DB industry perceives a particular risk to be more important to project success during the execution phase, and will reflect that perception in the size of the geotechnical contingency in their bid. If those risks 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 0.00 5.00 10.00 15.00 20.00 Ri sk A llo ca tio n [% ] ∆ Importance Index Shared DB Contractor 1 2 3 4 5 6 7 8 10 9 109 876 5 4 32 1

108 are shared in some manner, the DOT could potentially benefit from reduced price contingencies, mitigating the risk of not awarding in timely manner because all proposals are over budget or the engineer’s estimate. 3.3.2.8 Geotechnical Risk Expert Survey Findings Based on the survey findings and the comparison between the perceptions of DOTs and industry professional ranking of DB geotechnical risks, the following conclusions are drawn: 1. Twenty-seven DB geotechnical risk factors were identified by the analysis of the literature review, case studies, and expert geotechnical engineers. 2. Water table and settlement issues are the most frequent, highest impact, and most important geotechnical risks that should be addressed in a DB RFP. 3. Contaminated materials and soft compressible soils are the 3rd and 4th ranked geotechnical risks. 4. Table 3.18 can be used as a checklist during DB project development to ensure that geotechnical risk is adequately address in the DB RFP. 5. The perception of geotechnical risk factors between DOTs and industry differ in most of the 27 analyzed factors. Both descriptive and inferential analysis show that industry respondents tend to be more conservative in their perception of the relative importance of geotechnical risk factors than DOTs respondents. 6. The research results in Table 3.20 found similar perspectives on those geotechnical risk factors that DB delivery is not optimum a given project are: a. Landslides b. Subsidence (subsurface voids)

109 c. Contaminated material d. Prediction of subsurface conditions due to inaccessible drilling locations e. Sensitiveness of public considerations (Parks, historic building, etc.) f. Karst formations g. Slope instability The study recommends that these factors should be addressed in the RFPs and suggests they should optimally be shared between DOTs and industry in order to decrease potential contingencies in proposals. 7. Results show that DOTs currently shed nine of the ten top geotechnical risks. Thus, geotechnical risk sharing provides an opportunity to manage geotechnical uncertainties with an expected decrease in the amount of proposal price contingencies. The results support sharing geotechnical risk as a viable means to manage geotechnical uncertainty during the procurement phase. Finally, the findings of this study can be used to mitigate geotechnical risk by focusing the conduct of specific underground preliminary investigations on critical geotechnical risk factors. 3.4 Summary and Analysis of Case Study Interviews As part of Task 3a, a case study protocol was developed based on the findings from the surveys, the content analysis and the literature review with the objective of identifying, analyzing and understanding the current models for successful geotechnical risk management on projects delivered using DB project delivery. This protocol is then used to conduct structured face-to-face interviews, producing results with established points of comparison and a high level of reliability. Interviews were performed on eleven (11) case study projects identified by the research team across nine (9) states: California, Indiana, Maryland, Missouri, North Carolina, Ohio, South

110 Carolina, Texas and Utah. All of these case studies are projects that involved significant geotechnical risks and several differences and similarities were found in the way the DOTs managed them, which provided valuable insight towards identifying different strategies, tools and methods that are currently being used to handle the geotechnical risk. Table 3.22 shows a summary of all the case studies performed in the research. Table 3.22 Case Studies and Interviews Case No. Agency Interview Date Case Study Project Contract Award Year Contract Amount 1 Missouri DOT Feb-16 I-64 Daniel Boone Bridge over the Missouri River 2012 $111,000,000 2 Caltrans May-16 I-15/I-215 Interchange at Devore 2012 $208,000,000 3 Indiana DOT Jun-16 PR 69 from Taylor Ridge Road to 1435 west of CR 750E 2013 $110,000,000 4 Utah DOT Jun-16 I-15 Corridor Reconstruction Project 1997 $1,600,000,000 5 Utah DOT Jun-16 SR-73 Pioneer Crossing, Lehi 2009 $282,361,000 6 Texas DOT Jul-16 Dallas, Horseshoe Project 2012 $818,000,000 7 Maryland DOT Jul-16 IS-270 Innovative Congestion Management Project ~2017 $100,000,000 8 Ohio DOT Jul-16 Columbus Crossroad - Project 1 2011 $200,350,000 9 Ohio DOT Jul-16 Cleveland Innerbelt CCG1 (I90WB Bridge) 2010 $287,400,000 10 South Carolina DOT Aug-16 Port Access Road 2016 $220,700,475 11 North Carolina DOT Aug-16 I-40 Landslide Project 2004 $10,574,740 This section details the findings from the case study interviews and provides supporting information on the way the geotechnical risks were contractually handled by the different agencies. Table 3.23 shows a summary of the distinct characteristics representing how the risk was managed in terms of mitigation actions implemented pre-award.

111 Table 3.23 Summary of Case Studies–Geotechnical Risk Mitigation Actions Case No. State Case Study Mitigation Action 1 Missouri I-64 Daniel Boone Provide as much information as possible to proposers, accepting requests for additional investigation. 2 California I-15/I-215 Devore Perform additional studies if the project requires it. Prescriptive requirements. 3 Indiana PR 69 Perform as much geotechnical investigation as DBB. Mandatory soil improvement design. 4 Utah I-15 Reconstruction GDR nearly 100% AASHTO. Increase settlement warranty requirement. 5 SR-73 Pioneer Exclude owner's studies from DSC clause. Mandatory specifications and extended warranty for settlement. 6 Texas Dallas Horseshoe DSC clause with caps. Two NTPs. Not limiting contractor's geotechnical investigation. Mandatory pavement designs. 7 Maryland IS-270 Innovative C Progressive DB. Scope validation period. Appropriate amount of studies (not too much). 8 Ohio Columbus Crossroad First $250,000 of DSC is considered incidental. Requirement for contractor to provide GBR for tunneling. 9 Cleveland Innerbelt Robust subsurface exploration. Requirements for deep foundations and drilled shafts. A $500,000 DSC threshold was included. 10 South Carolina Port Access Road Extensive geotechnical and environmental investigation (including deep boring). Threshold for Hazmat. Seismic parameters provided. Elimination of geotechnical DSC. 11 North Carolina I-40 Landslide Provide as much raw data as possible in the contract. Allow requests for more borings. No DSC clause. Nested DB. The details of each case study project are contained in Appendix B. 3.5 Summary and Analysis of Legal Review The legal review analyzed the following major issues that impact the agency’s ability to both manage and mitigate geotechnical risk in DB projects:  Differing site conditions  Geotechnical liability issues on DB projects  Responsibility for geotechnical design decisions on DB projects  Contractual geotechnical risk tools. The complete details of the analysis are found in Appendix E. The important findings of the legal review are as follows:

112  The courts generally hold the owner in a DB project liable for differing site conditions regardless of the presence of exculpatory language that attempts to shift the differing site conditions risk to the design-builder. In those cases where the owner prevailed, it was generally due to a technicality such as untimely notification rather than contract risk shedding verbiage.  Absent a DSC, the liability for differing site conditions is theoretically shifted to the contractor, who will bid the DB project accordingly. However, given the result of the previous finding regarding the efficacy of exculpatory clauses, if a differing site condition is realized, the owner rarely prevails in the courts. Therefore, the inclusion of a DSC in a DB project was found to be an effective practice for mitigating geotechnical risk.  Three contractual tools were identified as effective geotechnical risk management approaches. All three provide a mechanism to negotiate the actual distribution of geotechnical risk after award based on a full geotechnical engineering investigation. 1. Scope validation period 2. Multiple notices to proceed 3. Progressive DB 3.6 Application of the Findings The analysis of DOT versus industry expert perceptions of geotechnical risk in DB projects demonstrated the need to align differences in perception of the geotechnical risk between owner and contractor to avoid over- or under-estimating the risks by either party and reduce conflicts by effectively sharing the risk. This is done by establishing a set of geotechnical risk management strategies that are then implemented using risk management methods and mitigation tools. This

113 section will synthesize the effective practices (tools) identified in the research within the context of available strategies and methods for managing geotechnical risk for DB projects. The major factors for which a risk mitigation strategy is needed to resolve common geotechnical issues present in most DB projects is shown in the list below. This list is not intended to be comprehensive, but rather to be general enough to apply to most, if not all, DB project delivery environments.  Delays due to untimely actions by third party stakeholders.  Inefficiencies in the project delivery process due to failure to include salient geotechnical risk issues in the procurement process.  Lost opportunities to avoid difficult geotechnical conditions.  Claims due to differing site conditions.  Poor quality post-award geotechnical investigations. The corresponding list of respective strategies to align differences in perception of the geotechnical risk between owner and contractor and address the causes of geotechnical-related issues on DB projects is as follows:  Implement early contractor design involvement through encouraging geotechnical ATCs during procurement.  Use the DB process to address other geotechnical involving third party stakeholders as early as practical in project development and delivery.  Raise the visibility of geotechnical issues in DB projects to ensure competing DB teams understand the level of criticality on each project.  Avoid differing site conditions claims through enhanced contract mechanisms designed specifically for addressing geotechnical risks.

114  Promote an atmosphere of life cycle-based design and construction decision-making with respect to geotechnical risk on DB projects. The tools identified in the research can be implemented using the specific methods for the purpose of executing the five risk mitigation strategies, as presented in this section. Table 3.24 through 3.28 matrix methods and tools necessary for implementing each of the five strategies identified in the Phase research. The source column lists at least one source from the literature, surveys, content analysis, and case studies. A two-letter state postal abbreviation notation shows the specific DOT where the tool was found in the surveys, content analysis, and case studies. A notation of state abbreviation plus (XX+) indicates the tool was found to be in use by more than one DOT. The entries in the source column are not intended to be exhaustive, but rather to provide the evidence of where the specifics tools were found to be effective. Table 3.24 Implement Early Contractor Design Involvement Method Tool Source Pre-advertising  Include GBR-C provision  Provide a mechanism to conduct competing team requested additional borings, i.e. permits, rights of access, etc.  Collect potential contaminated material information during ROW acquisition  OH  FL+  Literature Procurement  Request geotechnical ATCs on DB projects  Define no-go zones for geotechnical ATCs  Competitor designated boring locations  Competitors permitted to conduct supplementary borings at own expense  Progressive DB  Unit prices for contaminated material, over-excavation, etc.  WA+  CA+  UT+  MN+  MD  MT+ Preconstruction  Scope validation period  Multiple NTPs with one designated for geotechnical investigation, design, and a second specifically to commence excavations, utility work, etc.  Contractor produced GBR-C  Negotiated GBR interpretation  VA  MO  OH  WA+ Construction  Not applicable

115 Table 3.25 Involve Third Party Stakeholders as Early as Practical Method Tool Source Pre-advertising  Flexible footprint for NEPA clearance  Site conditions history from property owners during ROW acquisition  MO  Literature Procurement  Request of utility-related ATCs  GA Preconstruction  Assign design-builder responsibility for utility coordination  Multiple NTPs with one designated for geotechnical investigation, design, and a second specifically to commence excavations, utility work, etc.  TX+  Literature Construction  Multiple NTPs with one designated for geotechnical investigation, design, and a second specifically to commence excavations, utility work, etc.  Literature Table 3.26 Raise the Visibility of Geotechnical Issues Method Tool Source Pre-advertising  Geotechnical conditions database  Furnish GBR  Include GBR-C provision  Provide a mechanism to conduct competing team requested additional borings, i.e. permits, rights of access, etc.  Flexible footprint for NEPA clearance  Site conditions history from property owners during ROW acquisition  Performance specifications for post-construction performance (subsidence, etc.)  VA  WA+  WA+  OH  MO  Literature  MN+ Procurement  Include differing site conditions clause  Request of geotechnical ATCs  Define no-go zones for geotechnical ATCs  Competitor designated boring locations  Competitors permitted to conduct supplementary borings at own expense  Progressive DB  Unit prices for contaminated material, over-excavation, etc.  Weight geotechnical evaluation criteria  SC+  CA+  UT+  MN+  MD  MT+  VT+  MI+ Preconstruction  Scope validation period  Multiple NTPs with one designated for geotechnical investigation, design, and a second specifically to commence excavations, utility work, etc.  Contractor produced GBR-C  Negotiated GBR interpretation  VA  Literature  OH  WA Construction  Differing site conditions allowance  Contaminated material allowance  WA+  .MN+

116 Table 3.27 Enhanced DB Geotechnical Contract Mechanisms Method Tool Source Pre-advertising  Geotechnical conditions database  Furnish GBR  Include GBR-C provision  Prescriptive geotechnical design  Performance specifications for post-construction performance (subsidence, etc.)  VA  WA+  OH  OR  DE+ Procurement  Request of geotechnical ATCs  Define no-go zones for geotechnical ATCs  Competitor designated boring locations  Competitors permitted to conduct supplementary borings at own expense  Unit prices for contaminated material, over-excavation, etc.  SC+  MO+  HI+  MD  MT+ Preconstruction  Assign design-builder responsibility for utility coordination  Contractor produced GBR-C  Negotiated GBR interpretation  AZ+  WA  OH Construction  Differing site conditions allowance  Contaminated material allowance  WA+  MN Table 3.28 Life Cycle-based Design and Construction Decision-making Method Tool Source Pre-advertising  Geotechnical conditions database  Flexible footprint for NEPA clearance  Site conditions history from property owners during ROW acquisition  Performance specifications for post-construction performance (subsidence, etc.)  VA  MO  Literature  VA+ Procurement  Request of geotechnical ATCs  Include life cycle criteria in best value award scheme  UT+  TX+ Preconstruction  Validate proposed life cycle elements during design  Literature Construction  Encourage life cycle related value engineering proposals from subcontractors  Literature To summarize the application of research findings, aligning differences in perception of the geotechnical risk between owner and the design-builder can potentially avoid geotechnical risk issues, such as unnecessary contingencies and differing site conditions claims. The ultimate goal is to mitigate the geotechnical risks by applying strategies and tools in each stage of the DB delivery process that encourage collaboration between the parties in a DB contract.