Influence of Geotechnical Investigation and Subsurface Conditions on Claims, Change Orders, and Overruns (2016)

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

7 chapter two LITERATURE REVIEW Five topics of interest to the role of subsurface investigation on claims, change orders, and cost overruns were reviewed and provide the outline for this chapter. • Transportation agency standards for subsurface investigation. • A research report on geotechnical change orders at Indi- ana DOT. • The effect of subsurface investigation on claims, change orders, and cost overruns. • The human effect on claims, change orders, and cost overruns attributed to subsurface conditions. • The effect of contracting practices (e.g., design-build project delivery mechanism) on claims, change orders, and cost overruns attributed to subsurface conditions. TRANSPORTATION AGENCY SUBSURFACE INVESTIGATION PRACTICES U.S. transportation agencies have varying requirements and practices for subsurface investigation. Details of agency requirements and practice are presented in chapters three and four. This section summarizes three sources of national guid- ance regarding subsurface investigation that have informed many of the state agency guidelines. This section also pre- sents agency practices for performing subsurface investiga- tion, especially with respect to in-house investigation versus investigation through subcontracting. AASHTO Manual on Subsurface Investigations and LRFD Bridge Design Specifications The AASHTO Manual on Subsurface Investigations (AASHTO 1988) presents information and recommendations related to site characterization for all types of transportation facilities. A major revision of the manual is currently underway. The manual’s general recommendations, especially those for bor- ing spacing and boring depth (Section 7), have been adopted by many state agencies. The recommendations for boring spacing and boring depth were also adopted in the Founda- tions Section (Section 10) of the AASHTO LRFD Bridge Design Specifications (AASHTO 2014). National Highway Institute Manual on Subsurface Investigations FHWA’s National Highway Institute (NHI) offers a training course regarding subsurface investigations; the course man- ual (Mayne et al. 2001) is also referenced by the AASHTO LRFD Bridge Design Specifications (AASHTO 2014). The NHI manual includes similar information to that in the AASHTO Manual on Subsurface Investigations, but with significant updates resulting from advancements in technol- ogy and practice, especially those related to the cone penetra- tion test (CPT) and geophysics. FHWA Geotechnical Engineering Circular No. 5 FHWA published another manual that includes information regarding subsurface investigation, Geotechnical Engineering Circular No. 5 (GEC5): Evaluation of Soil and Rock Prop- erties (Sabatini et al. 2002). As with the AASHTO and NHI subsurface investigation manuals, GEC5 includes information on planning subsurface investigations; however, GEC5 also devotes significant attention to interpretation of subsurface investigation data for design purposes. The AASHTO LRFD Bridge Design Specifications (AASHTO 2014) also refers to GEC5 for information regarding subsurface investigations. A significant revision of GEC5 is underway. Agency Subsurface Investigation Capabilities Badger (2015) surveyed 36 state transportation agencies regarding agency practices for subsurface investigation. Most agencies (30 of 36) have in-house capabilities; the remaining six contract all exploration services. Half of the surveyed agencies reported their field exploration program had decreased over the previous ten years. One respondent noted in-house capabilities were more common for small projects, which had been associated with relatively high administrative costs for processing contracts and payments for external subsurface investigations. Another respondent noted all in-house capabilities were eliminated in 2005 because of the need for equipment replacement and limited resources. One respondent whose agency now predominantly uses contract drilling instead of in-house capabilities noted less drilling is accomplished per project because the contract drilling cost is greater. A 2007 NCHRP Synthesis survey of U.S. state and Cana- dian provincial transportation agencies found that nearly three- quarters of responding agencies used CPT on 10% or fewer of projects (Mayne 2007). Almost two-thirds of the respondents cited subsurface materials that were too hard to penetrate with CPT as an obstacle to its use. However, the same study found

8 that 64% of agencies anticipated an increase in their use of CPT. Indeed, Badger’s survey (2015) found that three- quarters of the 36 responding agencies used CPT, although the responses did not indicate how frequently each agency employed it. GEOTECHNICAL CHANGE ORDERS AT INDIANA DEPARTMENT OF TRANSPORTATION The causes and costs of geotechnical change orders at the Indiana DOT (INDOT) were studied by Prezzi et al. (2011) and Khan (2014) with objectives similar to those established in chapter one. The results of that project are summarized here to identify common sources of geotechnical change orders and their costs before those topics are explored more gener- ally for all agencies in chapters three and four. Prezzi et al. (2011) studied INDOT change orders associ- ated with work done by the agency’s geotechnical office over a 5-year period beginning in 2003. The work was motivated by an agency perception that change orders “attributed to geo- technical conditions” were “excessive” and perhaps increas- ing; the research was designed to quantify the number and cost of geotechnical change orders and to develop guidance for reducing them. The study included three components: 1. A national survey similar to that conducted for this synthesis. 2. Analysis of change order information from the ten largest contracts per year in each of INDOT’s six dis- tricts (300 contracts total). 3. Thirteen interviews with agency project engineers and external consulting engineers familiar with INDOT projects and practices. The national survey response rate was low and the survey results that were received were limited. Prezzi et al. (2011) focused primarily on results of the quantitative change order analysis. Several of the project’s results are most pertinent to this synthesis topic: • Quantitative analysis of change order data is compli- cated by difficulties associated with interpreting a large database of unique incidents that must be categorized by agency definitions. The authors chose to consider both “soil-related works” change orders; that is, all change orders associated with any construction activities or materials associated with geotechnical work (e.g., debris removal), as well as a more specific class of change orders with geotechnical causes, which were determined based on INDOT database reason codes; for example, Constructability: Soils Related. It was noted that the latter definition was more meaningful for geotechnical work because the former included change orders not directly related to geotechnical work. • The average cost of geotechnical change orders was 1.3% of the estimated total construction costs. • The cost of geotechnical change orders was just over 10% of the total cost of all change orders. • Approximately one-quarter of the projects (84 of 300) included geotechnical change orders, with many of these projects including more than one geotechnical change order. The project engineer and external consulting engineer interviews also produced information relevant to this synthe- sis. Four main causes for geotechnical claims based on the interviews were summarized, although some of the causes are associated more with design issues than with investigation problems: • Failure to identify poor subgrade that was frequently attributed to inadequate site investigation, but also resulted from improper plan elevations. • Pile overruns and underruns, which occur when the as-built driven pile depths are different from those shown on plans. • Erosion control material quantity errors often associ- ated with underestimating riprap and geotextile quan- tities as a result of mischaracterizing the soil drainage conditions. • Mechanically Stabilized Earth wall construction, although the changes were mostly related to no geotechnical aspects such as wall geometry conflicting with surface drainage lines. The interviewees also provided the following recommen- dations for reducing geotechnical change orders: • More boreholes as well as more flexibility in planning subsurface investigations considering geology, prior site knowledge, and region. • A design checklist addressing issues commonly encoun- tered during construction. • Expedient decisions when construction issues are encoun- tered because time was perceived to dramatically increase the cost of change orders. INDOT continued to track geotechnical change orders following the publication of the Prezzi report. In a presenta- tion to the FHWA Midwest Geotechnical Conference, Khan (2014) presented data from 2009 to 2013 showing that the average annual total cost for geotechnical change orders was $10 million, approximately 17% of all change orders and just less than 1% of the total amount the agency spent on con- struction. Khan cited inadequate geotechnical investigation as a primary cause of geotechnical change orders. He also mentioned many of the factors from the Prezzi (2011) report listed earlier. More details of INDOT’s experience with geo- technical change orders are presented in chapter four.

9 EFFECT OF SUBSURFACE INVESTIGATION ON CLAIMS, CHANGE ORDERS, AND COST OVERRUNS Gould (1995) provided useful information regarding how geotechnical construction risks are affected by subsurface investigation. Consistent with court and construction indus- try terminology, Gould defined two types of differing site condition claims, Type I and Type II. Type I changes refer to differences between encountered site conditions and those shown in contract documents, whereas Type II changes refer to a “surprise to all, not a discrepancy in the documents but an unusual physical condition beyond that reasonably expected.” Gould noted that while additional subsurface exploration reduces the risk of Type II changes, it can increase the risk of Type I changes by “offering a larger target to an aggrieved contractor.” Gould also identified four causes of changes that can occur even when a competent subsurface investigation is completed: 1. Surprise claims for incidents that defy experience with local geology and/or the construction task. 2. Incidents resulting from conditions that cannot be defined adequately from ordinary investigation meth- ods; for example, Gould cites underpredicting the size of a boulder. 3. Claims resulting from properties that are misunder- stood as a result of “limitation in the state of the art.” 4. Incidents involving features that are too small to be discovered given the precision of the subsurface investigation. Gould also provided detailed guidance for subsurface exploration practices to control claims. The guidance is pre- sented as 11 specific practice suggestions, which are intro- duced by a warning that “exploration focused too narrowly on design may be insufficient for construction.” Mott MacDonald and Soil Mechanics, Ltd. (1994) studied the effect of subsurface investigation on construction cost overruns by examining results from a database of 58 trans- portation projects in the United Kingdom. Three-quarters of the projects had cost overruns greater than 10% of the con- tract value. The authors reported “about half” of the overruns resulted from geotechnical causes, the most common being (1) problems from seepage and groundwater, (2) encounter- ing materials different in classification from those anticipated, and (3) removal and replacement of additional unsuitable material. The direct geotechnical cost overruns averaged 3% of contract cost, which the authors compared with an aver- age of 1% of contract cost spent on site investigation. Indi- rect claims resulting from delay and disruption were more significant, amounting to 5% of contract cost. It was noted that while most of the direct costs would have been required even with an adequate site investigation, the indirect overruns could have been avoided. The Mott MacDonald and Soil Mechanics, Ltd. 1994 report includes a graph of the total increase in final construc- tion cost versus site investigation cost, with both quantities expressed as a percentage of the project award cost (Fig- ure 1). The authors of this report refer to an “outer bound,” although the upper bound included in Figure 1 was added by Clayton (2001). The shape of the upper bound line is strong evidence that increased subsurface investigation does indeed reduce risks associated with geotechnical construction. How- ever, there are a significant number of projects with relatively Upper Bound? FIGURE 1 Graph of increases in construction cost for infrastructure projects as a function of cost of subsurface investigation (adapted from Clayton 2001 and Mott MacDonald and Soil Mechanics Ltd. 1994).

10 limited subsurface investigation costs that experienced only modest cost overruns. These projects, and, in general, the ver- tical variation of the data points for a given level of subsur- face investigation suggest there are other factors that affect the amount of the cost overrun. Part of the cost overrun varia- tion results from some ground problems being more difficult to resolve than others; however, it is also likely that “better” subsurface investigation practices produce better site charac- terization and result in more useful information for design and construction. A similar study was undertaken by the U.S. National Com- mittee on Tunneling Technology (USNCTT), which studied the effect of geotechnical site investigation on construction changes and claims. USNCTT described differing site condi- tion change orders and claims as “many” and “costly” (U.S. National Committee on Tunneling Technology 1984). Indeed, Gould (1995) summarized the data from the USNCTT study as including claims that amounted to 12% of the overall con- struction costs. The USNCTT study included 87 major tun- neling projects constructed over a 20-year period. USNCTT examined the ratio of completed cost to engineer’s estimate versus subsurface exploration quantity and cost data, which were available for 36 of the projects. The resulting plots reveal significant scatter, but USNCTT noted that engineer’s estimates become more reliable as the subsurface exploration quantity and cost increase. USNCTT recommends 1.5 linear feet of borehole per route foot of tunnel; according to the study, the cost of such an investigation is roughly equivalent to 3% of construction cost. Finally, improved subsurface investigation has other ben- efits for infrastructure projects. Many studies have noted that improved subsurface investigation results in design efficien- cies as well (e.g., Hoek and Palmeiri 1998; Clayton 2001; Ching et al. 2014). HUMAN EFFECTS ON SUBSURFACE CONDITIONS CLAIMS, CHANGE ORDERS, AND COST OVERRUNS Interestingly, several studies have concluded that geotech- nical risks are not exclusively attributable to ground condi- tions, but also involve human contributions. Based on the collective evaluation of several studies of geotechnical risks, Baynes (2010) concluded that “available information sug- gests that the ground conditions and the project staff respon- sible for the geo-engineering process are both significant sources of geotechnical risk and that the project staff may actually be the largest source.” Clayton (2001) described this “human” aspect of geotechnical risk as follows: There are numerous ways in which the ground can cause prob- lems for construction, for example due to chemical attack, heave, subsidence, groundwater flow, slope failure, excessive foundation settlement, and so on. Because of the considerable range of risks the ground can pose, it is relatively easy for an inexperienced or non-specialist designer, perhaps using routine procedures, to fail to recognise a critical mechanism of damage or failure that may threaten either the financial viability or health and safety of a project. If a mechanism of damage (a limit state) is not foreseen then it cannot be designed for, and it is often for this reason that ground-related problems occur. Similar conclusions were reached by Moorehouse and Millet (1994) in an analysis of 37 geotechnical consulting cases involving failure, which the authors defined as “the results of the unfulfillment of a claim, promise, request, need, or expectation between and among any of the design and construction parties and the client.” The study focused on engineering consulting services, but reached conclusions applicable to this synthesis. The second most common cause of failure, noted in 15 of the 37 cases, was “lack of disclo- sure of risks, uncertainties, and consequences,” meaning the engineer failed to effectively advise owners or contractors about geotechnical risks that ultimately came to fruition. The most common cause of failure was “recommendation not fol- lowed by client or contractor,” which has similar albeit more obvious roots in human error. The authors’ recommendations emphasize the responsibility of management personnel to staff and train technical personnel appropriately and to “deal with the real need for intelligent disclosure of risks, uncertain- ties, and consequences.” The findings of Baynes (2010), Clayton (2001), and Moorehouse and Millet (1994) suggest that human effects are a primary cause of subsurface conditions claims, change orders, and cost overruns, likely equal in importance to the more tangible effects of geotechnical investigation and con- struction practices. Clear and transparent communication of risks, assumptions, expectations, and consequences among agency, contractor, and consulting personnel is likely critical to reducing subsurface conditions claims, change orders, and cost overruns. EFFECT OF CONTRACTING PRACTICES ON SUBSURFACE CONDITIONS CLAIMS, CHANGE ORDERS, AND COST OVERRUNS Even a quick reading of literature related to claims, change orders, and cost overruns attributed to subsurface conditions reveals that contractual issues play a significant role. Con- struction contracts allocate risks between owner and builder. Typically, subsurface risks are allocated to owners through a differing site condition clause. Contractual issues are not a focus of this synthesis; however, two contract topics—bid documents and design-build arrangements—are summarized here because of their relevance to the synthesis topic. Geotechnical Bid Documents: Lessons from the Tunneling Industry The high frequency of litigation encountered in tunneling practice has motivated the tunneling industry to advocate practices that reduce litigation. Gould (1995) summarized

11 recommendations by the Underground Technology Research Council (UTRC) as follows: • Include a changed-condition clause in the construction contract. • Disclose fully in contract documents all geological data and interpretation. • Eliminate disclaimers that discount the value of the included geotechnical data. • Escrow bidding documents that show assumptions and pricing of the successful bidder. • Include a geotechnical design summary report in con- tract documents. • Establish in the contract a dispute review board for expeditious review and settlement of disputes. The last recommendation, establishing a dispute review board (DRB), has gained favor in the tunneling profession since UTRC made its recommendation. The Eisenhower Tunnel in Colorado, constructed in the late 1970s, was one of the first projects to include a DRB; by 1990, the use of such boards was prevalent, with the cost of projects using DRBs amounting to 70% of all tunneling project costs (Gould 1995). Gould noted that schedule advantages of DRBs exceed simply avoiding delays associated with the court system: DRB mem- bers are familiar with the project background and progress; therefore, disputes can be resolved quickly using on-site per- sonnel. UTRC estimated that the costs of maintaining a DRB ranges from 0.1% to 0.3% of total project construction cost. The geotechnical design summary report is another of the UTRC recommendations that has been implemented, although its name has evolved to Geotechnical Baseline Report (GBR). Recommendations for GBRs were outlined by ASCE (2007) in an update to the original recommendations of the UTRC. The updates were primarily motivated by a desire to expand the application of GBRs beyond tunneling to include deep foundations and highways, among other types of construction. ASCE’s Suggested Guidelines identify the GBR as the “single interpretative report” to be included with bid documents; all other factual information from subsurface investigations are to be included as a Geotechnical Data Report; however, it essen- tial that the GBR prevail over the Geotechnical Data Report. The baseline of GBRs refers to the anticipated site conditions presented in the report. Contractors assume risks associated with conditions consistent with or more favorable than the baseline; owners assume risks associated with conditions less favorable than the baseline. The ASCE Suggested Guidelines provide additional recommendations for establishing base- lines, including for design-build projects. Design-Build Contracts The 21st century has seen a significant increase in design-build contracts used for transportation construction, and FHWA featured design-build as a 2012 Every Day Counts initiative. Among many perceived benefits of design-build (e.g., acceler- ated project schedules, promotion of innovation, and efficien- cies related to having designers and constructors on one team), one is especially relevant to this synthesis: the opportunity to reduce owner risks compared with arrangements typically assumed for design-bid-build contracts. This opportunity is primarily manifested by eliminating the risk of claims related to design errors; however, agencies may also shift some of the risks of subsurface conditions, typically assumed wholly by the agency, to the design-builder (Gransberg and Loulakis 2012). Gransberg and Loulakis present a thorough review of con- tractual practices for geotechnical aspects of design-build projects in NCHRP Synthesis 429: Geotechnical Information Practices in Design-Build Projects (2012), which empha- sizes the accelerated schedule of design-build projects, with contracts frequently awarded before subsurface investigation is complete, as requiring especially competent management of geotechnical risks. This synthesis cites four measures commonly employed by agencies to manage geotechnical risks on design-build projects: 1. Selecting only design-build teams with significant geotechnical experience. 2. Assigning the most qualified agency geotechnical per- sonnel to design-build project oversight. 3. Limiting geotechnical designs to those in which the agency is confident. 4. Retaining quality management roles and responsibili- ties for geotechnical features in house. This report devotes significant attention to the unique contractual issues associated with design-build projects. Frequently, the design-builder will have at least some if not substantial responsibility for developing and performing the subsurface investigation. In response, agencies have shifted some of the contractual risk for differing site conditions to the design-builder; however, 10 of 11 design-build contrac- tors interviewed for the synthesis indicated that the ambigu- ity regarding how these shifts would be implemented was cause for concern. Only one agency, Washington State DOT (WSDOT), was identified by a contractor as having an unam- biguous contract provision. WSDOT’s “risk sharing clause” establishes a threshold dollar amount for differing site con- ditions for which design-builders assume risk; above the threshold, WSDOT assumes differing site condition risks. Another contracting practice identified as effective was establishing a process for “expeditious resolution of discrep- ancies between pre-award and post-award geotechnical con- ditions.” Both the expeditious review process and the risk sharing clause are similar to practices employed in the tun- neling industry. In his discussion of geotechnical risks and human factors, Clayton (2001) also alluded to design-build effects. He stated that increasing use of innovative contracting methods such

12 as design-build, which “disperse design responsibility,” is likely to increase ground-related problems unless changes are made to subsurface investigation practices. Clayton’s prediction is consistent with the observations of Grans- berg and Loulakis’s (2012) of increased geotechnical risks. Unclear, however, is how design-build risks are to be miti- gated and whether responsibility will be assigned to agencies or design-builders when they are not. SUMMARY OF SIGNIFICANT FINDINGS • AASHTO guidance regarding subsurface investigation is provided in the Manual on Subsurface Investigations (AASHTO 1988) and in the LRFD Bridge Design Speci- fications (AASHTO 2014). The recommendations in the AASHTO documents are consistent with one another. • A survey of 36 state transportation agencies by Badger (2015) indicated that agency in-house subsurface inves- tigation is common but decreasing, while use of the CPT is less common but increasing. Badger’s survey found that three-quarters of the 36 responding agencies (27) used CPT. • A study of INDOT change orders by Prezzi et al. (2011) found that quantitative analysis of change order data was complicated by database issues. • The study by Prezzi et al. (2011) also found that the aver- age cost of geotechnical change orders was 1.3% of the estimated total construction costs, that the cost of all geo- technical change orders was just over 10% of the total cost of all change orders, and that approximately one-quarter of the projects included geotechnical change orders. • Gould (1995) described two types of differing site con- dition claims: Type I, which applies to discrepancies between conditions depicted by contract documents and actual conditions, and Type II, which are a “sur- prise to all.” Additional site investigation reduces the risk of Type II claims, but may increase the risk of Type I changes. • Clayton (2001) analyzed data from a study by Mott MacDonald and Soil Mechanics, Ltd. (1994) and found that the magnitude of subsurface conditions cost over- runs decreases with increased site investigation expense relative to total construction cost. However, many proj- ects with relatively limited site investigations experi- enced only modest cost overruns. • The study by Mott MacDonald and Soil Mechanics, Ltd. (1994) found indirect costs associated with subsurface conditions claims, change orders, and cost overruns to be greater than direct costs. • Studies by Baynes (2010), Clayton (2001), and Moore- house and Millet (1994) suggest that human effects are a primary cause of subsurface conditions claims, change orders, and cost overruns. • Gould (1995) described recommendations from the tun- neling industry related to bid practices intended to reduce claims, change orders, and cost overruns. Most promi- nent were use of DRBs to resolve issues quickly and the creation of GBRs. GBRs are also recommended by ASCE (2007) because they provide a “single interpretive document.” • Geotechnical risks may be greater on design-build proj- ects. Gransberg and Loulakis (2012) summarized four recommendations for managing geotechnical risks on design-build projects, including selecting consulting teams with significant geotechnical qualifications and retaining quality management in-house.