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.

A -1 Appendix A: State of the Practice Introduction This section reports on the evaluation of current applications of geotechnical risk management practices using DB project delivery and the way these practices are applied on a variety of project types. For that purpose, several research instruments are utilized: extensive literature review, legal review of case law in the subject, content analysis of 59 issued letting documents from DOTs across the US, two nation-wide surveys to furnish real-time perceptional data regarding current practices managing geotechnical risks, and 11 case study interviews to collect detailed practices from completed DB projects with significant geotechnical risks. Brief Summary of Synthesis 402 Findings NCHRP Synthesis 429: Geotechnical Information Practices in Design-Build Projects, was the genesis of this research, identifying the need for DB geotechnical risk guidelines. The synthesis benchmarked the state of the practice with regard to the content of the geotechnical technical information found in DB solicitation documents. It concluded that the high-level federal encouragement via the Every Day Counts initiative (Mendez 2010) that state DOTs adopt DB project delivery made the need to thoughtfully manage geotechnical risk a critical project success factor. Previous research found that owners use DB delivery as a means for accelerating a project’s delivery (Songer and Molenaar 1996), and that the major barrier to expedited project delivery is the owner’s permission to release design product for construction (Higbee 2004). Geotechnical site investigation that drives the project’s design is usually the first task that must be completed. Geotechnical uncertainty is usually high at the time of DB contract award, putting the design- builder’s geotechnical team under pressure to produce a produce that can be released for construction so that site drainage, foundation and other subsurface construction can begin.

A -2 Therefore, “successfully managing the geotechnical risk in a DB project is imperative to achieving the requisite level of quality in the finished product.” (Gransberg and Loulakis 2011). Table A.1consolidates the effective practices reached NCHRP Synthesis 429 based the phase of project delivery in which they are found. Table A.1- Summary of Synthesis 429 Findings. Delivery Phase Effective Practices Literature Cite Procurement RFQ Require highly qualified geotechnical personnel Scott et al 2006 Assigning the agency’s most qualified geotechnical personnel to DB project oversight Potter and McMahon 2006 Tailor relative geotechnical weight for each DB project FDOT 2011, VDOT 2010, WSDOT 2004 Procurement Legal Explicit differing site conditions clauses Loulakis and Shean 1996 Risk sharing clauses that quantify the geotechnical risk a design-builder is exposed to with the DOT assuming everything above that threshold WSDOT 2004 Procurement ATCs Confidential one-on-one meetings to clarify request for proposal (RFP) intent and to present potential alternative technical concepts (ATC) Carpenter 2010 Utilizing confidential pre-approved ATCs to enhance innovation in geotechnical design and subsurface construction means and methods MnDOT 2003 Procurement Technical Mandating the use of geotechnical design solutions with which the agency is confident Papernik and Farkas 2011 Pre-proposal approval of geotechnical design approach UDOT 2005 Permitting design-builders to request/obtain additional site investigation prior to submitting a proposal WSDOT 2000 Allow competing design-builders to conduct their own pre-bid geotechnical investigations before developing their proposals UDOT 2005 Issue a draft DB RFP and ask for comments VTrans 2010 Require the geotechnical engineering design QA plan in proposal TTA 2001 Design Expedited geotechnical design review and acceptance procedures WSDOT 2004 Restricting the DOT to a single interim design review before final release for construction review Mississippi 2005 Maximizing the use of formal and informal over-the-shoulder design reviews MnDOT 2001 Permitting release of geotechnical design packages for construction before remainder of design is complete Arizona DOT 2001; Maine DOT 2010 Early design-builder-requested design reviews on the adequacy of the geotechnical design TTA 2001; NCDOT 2009 Construction Treat the geotechnical QM program differently than the remainder of the project by increased agency involvement Schaefer et al. 2001; Smith 2008 Early notice to proceed to conduct site investigation and physically locate utilities Hatem 2011 Use of selective unit pricing as done by the Montana, Delaware, and Virginia DOTs provides an effective means for managing geotechnical quantity risk MDT 2011

A -3 The major conclusions along with the primary literature source from NCHRP Synthesis 429 are as follows:  “DOTs typically select DB to accelerate project delivery (FHWA 2006).  DOT geotechnical design approval is a major hurdle to starting construction (Christensen and Meeker 2002).  Geotechnical uncertainty is always high until the post-award site investigation and geotechnical design report can be completed (Hatem 2011).  Geotechnical and site engineering is the first major design package and the one with the highest pre-award uncertainty (Higbee 2004). Therefore: o It must be completed as expeditiously as possible (Koch et al 2010). o The DOT needs to reduce the impact of geotechnical uncertainty as expeditiously as possible (Kim et al 2009).” (Gransberg and Loulakis 2011). Those findings can be consolidated and summarized in the following manner: DB geotechnical is best mitigated by making every effort to get to the point where the construction contractor can start excavations to physically identify actual site conditions. Geotechnical Risk in DB Projects DB vs DBB delivery In this section, we highlight the effect of project delivery (DB vs traditional DBB) on risk allocation among the parties to the contract. In DBB delivery system, the design is complete before the construction phase starts. The owner assumes the responsibility for accuracy of design. Much of the geotechnical claims falls under changes due to Differing Site Conditions (DSC). The FHWA mandates the use of a DSC clause for DBB projects on federal aid highway projects, unless its use

A -4 is contrary to the state law. The typical DSC clause provides broad relief to a contractor for site conditions that differ materially from what is expected according to the contract documents. So in DBB projects, the risk of differing site conditions is almost always borne by the owner (Tufenkjian 2007). In DB delivery however, the design-builder is responsible for completing the design and construction under a single contract. FHWA does not have a similar DSC mandate for DB projects, but encourages state DOTs to use DSC clauses when appropriate. The liability of the contracting parties due to the risk of differing site conditions is far from clear (Clark and Borst 2002; Loulakis et al 2016). In the absence of DSC clause, many DB contracts require that the design-builder conduct a comprehensive geotechnical study of the site. Some of these contracts, in the absence of DSC clause, include disclaimers of liability for the geotechnical information furnished by the owner during the procurement process due to the preliminary nature of the information and data. As a result, when the design-builder claims that it encountered a DSC based on owner-furnished geotechnical data, the owner may deny the claim because it contends the furnished information was preliminary and the design-builder should have done its own complete geotechnical assessment of the site. Many state DOTs however, use DSC clause in their DB contracts. A comprehensive survey of state DOTs conducted for this research showed that out of 24 respondents, only four states indicated that they do not use DSC clause in their DB contracts. A comprehensive description of this survey and analysis of contents is provided in this interim report. One important issue is how to establish an appropriate baseline for the DSC clause, if the owner decides to include such clause in the contract. The use of DB delivery system by public agencies goes back almost 50 years on building projects but it was not used in highway projects until the beginning of the 21st century (Gransberg et al 2016). One major reason was the difference in the scale of site conditions. While building site is limited to little more that the building footprint,

A -5 the site for highway project extends the length of the highway and may cover many different geotechnical conditions. The question in establishing the geotechnical baseline for DSC clause then becomes one of the scope and extent of geotechnical information furnished by the owner. Generally speaking, if the geotechnical information is limited, the design-builder may try to protect itself against geotechnical uncertainty by inclusion of higher contingencies in the contract price, especially in the case of fixed price contracts. Conducting a more comprehensive geotechnical investigation by the owner not only encourages the bidders to reduce their contingencies, it provides the owner more information about the project, its challenges, and expected costs. However, some owners may believe that providing more of the design data may somehow relieve the contractor of its obligation to do a comprehensive geotechnical assessment and to provide more opportunities for claims. Also, because in many DB projects, the speed of delivery is a major factor, extensive geotechnical investigation by the owner may delay project start time (Gunalan 2016). Accelerated procurement schedules may not allow the owner to research, collect, and share all the available geotechnical information to the bidders. So, in general, potential risks are created for both parties on a DB project that are not present in a DBB delivery system (WSDOT 2004). The owner is faced with the question of how much information and data is adequate for the geotechnical characterization and design of a highway, bridge or tunnel project. Hoek and Palmeiri (1998) present. Figure A. 1 is based on the data collected by the U.S. National Committee on Tunnel Technology (USNCTT 1984) on 84 tunnel projects. The plot shows the variations in cost against the ratio of borehole length to tunnel length. The outcome suggests that more extensive exploration and testing leads to reduced changes in the tunnel projects.

A -6 Figure A. 1 - Changes vs ratio of borehole length to tunnel length (Hoek and Palmeiri 1998) Cost of geotechnical investigation Needless to say, the extent of geotechnical investigation should be considered vis-a-vis the cost of this effort. While data on the cost of geotechnical investigations is scarce for DB projects, there are some suggestions for projects regardless of delivery method. Smith (1996) estimates the cost of ground investigation at less than 1% of construction costs. Van Staveren (2006) reports on several authors that estimate this cost to be in the range of 0.1% to 1%. The US Subcommittee on Geotechnical Site Investigations suggested a ground investigation budget of around 2% of total costs in order to keep the final cost within a margin of ±10% in tunneling projects. Clayton (2001) describes a study by Mott McDonald and Soil Mechanics Ltd (1994) that reported on the cost overruns as a function of expenditure on site investigation for UK highway projects (Figure A. 2). According to Clayton, less than 1% of project costs is expensed on ground investigations and it is

A -7 evident that spending more upfront helps to reduce the final cost of the project. The research effort concluded that the 1% level was often not sufficient to prevent large cost overruns in British highway projects. Figure A. 2 - Cost overrun vs site investigation expenditures (Mott McDonald & Soil Mechanics Ltd 1994) The research team interviewed nine state DOTs and collected data on the cost of geotechnical investigations for 11 DB projects. The range was reported to be between 0.25% to 0.5% of total project budget. Mostly, these efforts resulted in preparation of the Geotechnical Data Report (GDR) which contains the test data without detailed interpretation. Probably more important was the time impact. Duration of investigations varied between six months to 1.5 years depending on the size and scope of the project and the level of effort. A more detailed description of these interviews is provided in this Appendix C of the report.

A -8 Figure A. 3 provides a flowchart for the process of developing a geotechnical investigation plan as envisaged in this research. It is understood that extensive risk assessment may not be necessary for routine and simple projects. However, evidence shows that most of DB projects in transportation sector tend to be larger projects and their complexity is often increased because of the need to deliver them quickly. In the case of complex projects, risk assessment becomes an important step in the process. Its purpose is to identify major geotechnical risk factors and their magnitude. Based on the collective impact of geotechnical risks, a decision can be made regarding the extent of further investigation by the agency, allowing the bidder to request more testing conducted by the agency or by the bidder, or in rare cases, consider an alternative delivery method. Figure A. 3 - Development of geotechnical investigation program In cases where the level of risk is high, or the impact of full investigation on cost and schedule is prohibitive, and if the local laws permit, the use of progressive DB could be considered. This approach allows the contractor to proceed with the project and the final contract value will be determined as the scope becomes clear. However, many states may not allow any delivery method that is not based on a fixed price contract. Another related question is to establish the optimum level of geotechnical investigation from the owner’s point of view. If the risk assessment

A -9 process can establish the expected cost of unmitigated geotechnical risks and these costs can be compared to the benefits of reducing these risks given certain expenditure on testing and investigation, it would then be possible to establish the optimal level of geotechnical investigation. Observational Method The observational method in applied soil mechanics was first introduced by Peck (1969) and is based on utilizing field measurements and observations to validate or modify design assumptions during construction. Using this approach, the builder can start the construction phase assuming realistic (and not pessimistic) values for ground related parameters and depending on what will be unearthed during the construction phase, adjust the design values. Although the concept of design modification based on fresh information appears to be logical, the traditional DBB delivery poses obstacles in its implementation. The owner has to fully design the project without the benefit of what comes to light during construction and any deviation from the original design will be the responsibility of the owner and should be implemented through change orders. In DB delivery system, part of design and construction can be overlapped and the design-builder can use the observational method to identify, quantify and mitigate risks as they occur. In fact, observational method extends the design phase into the construction phase (Nossan 2006). Farnsworth (2016) summarizes the observational method into the following steps: (1) investigation of subsurface conditions, (2) assessing the most probable conditions and potential deviations, (3) designing for probable conditions, (4) selecting the required observations and calculate the anticipated values, (5) calculate the outcomes associated with each unfavorable condition, (6) selecting a mitigation action for each significant unfavorable event, (7) gathering actual observations during the construction phase, and (8) modifying the design as necessary. In other words, the observational

A -10 method follows the traditional steps in risk management, i.e., identification, measurement, and mitigation of risks. So while DB approach creates a challenge for the owner in terms of level of geotechnical investigations needed, it also provides some flexibility in terms of dealing with geotechnical risks and mitigating them. With this background on the distinctions between DBB and DB deliveries in the context of geotechnical risks, we now embark on describing various aspects of geotechnical risk management. Geotechnical Baseline In order to assist in administration of the DSC clause the owner needs to establish baseline statements in the contract documents. There are multitude of reports used in the industry to characterize and baseline geotechnical conditions. As an example NYSDOT uses two main documents for managing their DB projects: (1) Geotechnical Data Report (GDR) that contains all the factual geotechnical data collected for the project including the test results conducted by the agency and (2) Geotechnical Baseline Report (GBR) which is an interpretive document used for establishing “a common understanding between the contractor and the Owner of the subsurface conditions and their potential impact and effect of risk on the design and construction of the project design concept.” (NYSDOT 2013). A more detailed description of these reports are provided elsewhere in this interim report. The Technical Committee on Geotechnical Reports of the Underground Technology Research Council of the U.S.A. developed the concept of GBR in the 1990s (Essex 2007; Van Staveren 2006). The primary purpose of the GBR is to develop a set of contractual statements that describe geotechnical conditions anticipated during the construction phase. This becomes a risk allocation device because risks related to geotechnical conditions consistent or less adverse than the baseline values are allocated to contractor and risks resulting

A -11 from more adverse values defined by the baseline are allocated to the owner. Without the GBR or some other source for the design baseline it would be difficult if not impossible to assign risk responsibility to contract parties. In the extensive interviews conducted for this research, 11 projects in nine DOTs were investigated. In the majority of these projects, the agency only provided the GDR for characterization of geotechnical conditions. It could be due to the fact that the agency does not want to commit itself to interpretation of geotechnical conditions and feels that is the responsibility of the contractor. It seems however that in order to enforce a DSC clause presence of the GBR would facilitate the allocation of risk and reduce the contingency that design- builder may want to include in the price. Essex (2007) provides a procedure for using DB delivery method in dealing with geotechnical design issues. Figure A. 4 summarizes the proposed approach. The owner collects all the relevant data from similar previous projects and the test results for the current project (GDR). This information is made available to bidders and they are afforded the opportunity to request more information or tests which will be carried out by the agency if deemed appropriate, and made available to all bidders. The data is mainly gathered by the owner and interpreting the results and design decision-making is left to the bidder. Based on the data collected and test results the owner prepares a GBR for Bidding (GBR-B). The main emphasis of this document is the physical nature of the underground conditions, and much less the on the behavioral baselines that depend on equipment, and means and methods used by the contractor. The owner shall update sections of GBR-B during the bidding process to include the results of any additional testing carried out by owner based on bidders' request.

A -12 Figure A. 4 - Use of Geotechnical Baseline in DB projects The bidder will use the GBR-B and augment it by their approach and design to come up with a new version of the GBR for Construction (GBR-C). GBR-C will then serve as the baseline for allocating risks encountered during construction. It will incorporate the contractors’ interpretation and approach and its proposed means and methods. The owner should review the GBR-Cs prepared by bidders and seek clarification on any issues as needed. Based on this review, the owner will have the opportunity to seek adjustment of contract price if appropriate. After the bidder is selected, its GBR-C will be incorporated into the

A -13 DB contract documents and forms the basis for risk allocation during the design-construction phase. From the above discussion, it becomes evident that baselining the geotechnical conditions is a prerequisite for a fair and balanced risk allocation, especially in case a DSC clause is used in the DB contract. When a risk is encountered, in presence of a carefully developed GBR-C, it would be possible to assign the risk to the party who is responsible and to plan for mitigating the consequences of the risk event. Geotechnical Risk and Uncertainty Much of geotechnical risk is due to uncertainties in soil characteristics. Engineering properties of soils vary from one location to another. This variability becomes significant in transportation projects as they encompass large areas and long distances such that measuring soil properties and characterizing soil behavior will be fraught with uncertainties. Baecher (1987) and Christian and Baecher (2011) divide uncertainties of soil response to design loads into two categories: (1) data scatter and (2) systematic uncertainty (Figure A. 5). Data scatter is composed of actual spatial variability from one point to another in soil mass, and the noise introduced by measurement methods. Systematic uncertainty is due to statistical error because of limited number of observations and model error due to the approximate nature of mathematical models used for soil behavior prediction.

A -14 Figure A. 5 - Sources of error or uncertainty in soil property estimates (Baecher 1987) The traditional approach has been to use factors of safety to deal with these uncertainties. There are fundamental issues with the traditional approach which is beyond the scope of this research. One way to deal with geotechnical uncertainty is to use a risk-based design approach. In this approach, variability of design parameters is explicitly considered and using probabilistic approaches, the uncertainty in the outcome of the analysis is quantified. The designer can then decide on the design dimensions in such ways to meet desired confidence levels against various events. A conceptual example is shown in Figure A. 6. In evaluating the slope stability, the uncertainties in input values such as soil properties, water table level, and site geometry is modeled using probability distributions and the probability of failure for the slope is calculated. Based on a desired confidence limit (such as probability of failure lower than 0.001), the embankment can be designed. There is a cost to design conservatively and while using higher confidence levels will result in relatively safer structures, this cost should be evaluated against the benefits of such decisions. The uncertainties in geotechnical properties should ideally be translated into the

A -15 uncertainties affecting the project cost and duration. The current practice in designing routine earth structures encountered in highway work, does not involve probabilistic approaches. This is due to lack of sufficient data, complexity of approaches, level of effort required, and the unresolved theoretical issues (Christian and Baecher 2011). Figure A. 6 - A risk-based design approach for slope stability (Clayton 2001). However, it is important to keep in mind that these uncertainties may impact the cost and schedule of the project. Identifying, measuring, and mitigating risks arising from these uncertainties is the subject of geotechnical risk assessment discussed in the following sections. The Proposed Risk Approach Previous sections provided some background regarding geotechnical investigations and their related risks in design-build projects. Differences of DB delivery system vs traditional DBB were highlighted and the role of DSC clause, an augmented GRB, and the nature of geotechnical

A -16 uncertainties were described. This information sets the stage for posing the main question for this research. With design-build comes the distinct possibility that underground investigations will be undertaken by the contractor after the price of the project has been established. “Therein lies the question: how much investigation, if any, should the owner conduct prior to advertising the DB contract to characterize the geotechnical conditions upon which the competing design-build teams must base their price?” (Gransberg et al 2016a). The research team has developed a modeling approach summarized in the following flowchart ( Figure A. 7). Figure A. 7 - The proposed modeling approach for geotechnical risk analysis in DB projects

A -17 The proposed modeling approach follows the conventional risk assessment approach used in various state DOTs and transit agencies. The process has been described in several sources (Allen and Touran 2005; Molenaar 2010; D’Ignazio 2011) and can be summarized in three basic steps as depicted in Figure A. 8. Figure A. 8 - Risk Assessment Process Each of the three steps shown in Figure A. 8 can be broken down into further components. Here, our main emphasis is geotechnical risks and we have incorporated these three steps in the flowchart. Note that the whole process can be thought of as a loop. This means that mitigation effort can rearrange the risks in ways that identification and assessment of this rearranged risks may become necessary. Risk Identification Risk identification is the process of identifying risks that can adversely affect the project cost and schedule and also the opportunities that can reduce project costs or result in a reduction in project duration. In general, a comprehensive risk management approach encompasses all risks and opportunities that can affect a project. One suggested approach is to develop a Risk Breakdown 1. Risk  Identification 2. Risk  Quantification 3. Risk  Mitigation &  Monitoring

A -18 Structure (RBS) to categorize risks and to cross-reference risk factors to individuals responsible. RBS was first introduced by Hillson (2002) and offers a WBS-style format for structuring the risks. It organizes the risks into a number of source-oriented groupings in a hierarchical manner. The project risks at the highest most aggregate level could be divided into (1) External, (2) Organizational, (3) Project-specific, (4) Legal and contractual, etc. The complete risk assessment will cover all these areas in order to assess the project’s risk profile and arrive at reasonable mitigation measures. In this research, our emphasis is much more limited and focused on geotechnical risks that can be identified through site investigation and testing. As such, most of the types of risks that are of interest in this work fall under the category of Project-specific design- related technical risks. Risk identification should start at the very earliest phases of project development. Early identification of major project risks will help in avoiding the selection of problematic project solutions and costly remedial action in later phases. It is a well-established fact that as the project design gets further along, changes to project scope will become more costly. Considering various phases of project development (Figure A. 9), risk identification by owner’s team should start during the Conceptual Design phase. In most transportation projects, the owner tends to perform a fair amount of design before going to DB Contractor selection. Much of Preliminary Engineering is done by the owner. Needless to say, a comprehensive project risk assessment must be done by the end of preliminary design phase.

A -19 Figure A. 9 - Generic DB Timeline In terms of categories of geotechnical risks, Van Staveren (2006) suggests the following for the main elements of ground and their related risks: 1. Ground (soil, rock) – geotechnical risk 2. Groundwater – geohydrological risk 3. Contamination of soil or water – environmental risk 4. Manmade structures (such as buried piles, archeological sites) – manmade obstruction risk Another categorization could be based on the general area of geotechnical design effort. Geo-Institute of ASCE has divided geotechnical design into the following nine areas: 1. Dewatering 2. Deep foundations 3. Shallow foundations 4. Mass excavation and grading 5. Marine construction 6. Ground improvement 7. Slope instability 8. Anchored earth retention 9. Tunneling

A -20 Risks can be identified based on the areas listed above. There will be many ways the project team can categorize geotechnical risks. The purpose of having a reasonable system for risk breakdown is to ensure that no significant risk is disregarded. An effective method would be to develop a detailed Risk Catalog prepared to the desired format and use it on every project. As the time goes by, the agency will be able to prune the developed risk catalog by deleting irrelevant risk factors and add newly identifed risks to the list.