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.

13 chapter three SURVEY RESULTS A three-part survey was administered to 55 agencies, includ- ing state transportation agencies for all 50 states, Puerto Rico, and the District of Columbia, as well as the three divisions of the Office of Federal Lands Highway. The complete sur- vey questionnaire is provided in Appendix A, and complete survey results are presented in Appendix B. Part One of the survey contained questions related to subsurface investiga- tion practices. Part Two included general, mostly qualita- tive questions about claims, change orders, and cost overruns. Part Three requested specific quantitative information regard- ing claims, change orders, and cost overruns, and was optional because it was anticipated that the quantitative data requested would not be readily available for many agencies. The sur- vey was distributed to agency geotechnical engineers; how- ever, the survey instructions encouraged the engineers to share the survey with construction colleagues, especially for help with completing Parts Two and Three. Fifty-one agen- cies responded to the survey, including 46 state DOTs, which corresponds to a response rate of 92% for the state agen- cies. Responding agencies are shown in the map in Figure 2. Of the 51 responding agencies, 11 included at least a partial response to Part Three. This chapter summarizes the results of the survey, including subsurface investigation practices; qualitative and quantitative information regarding the nature of claims, change orders, and cost overruns attributed to sub- surface conditions; and the relationship between subsurface investigation practices and claims, change orders, and cost overruns. SUBSURFACE CONDITION PROBLEMS AND SUBSURFACE INVESTIGATION PRACTICES Several of the questions in Part One of the survey inquired generally about problems related to subsurface conditions the agencies might have experienced. As shown in Table 1, a majority of responding agencies (63%) stated they expe- rienced a modest number of problems resulting from sub- surface conditions, whereas only two agencies responded that they experienced frequent problems. Another question in Part One asked the respondent if his or her agency had experienced significant performance problems that could be attributed to subsurface conditions or site characterization practices. Forty- one percent of respondents indicated their agency had expe- rienced such problems, whereas 53% indicated they had not. Together, the responses to these two questions indicated that two-thirds of agencies experience problems attributed to subsurface conditions, and the severity and frequency of the problems vary. The remaining one-third has infrequent and relatively insignificant problems related to subsurface conditions. Responses from Table 1 are shown by agency in Fig- ure 3. There are some geographic trends that can perhaps be explained by geologic regions. About half of the agencies with infrequent and relatively insignificant problems related to subsurface conditions are located in the Central Lowlands and Great Plains geologic regions of the Midwest. The agen- cies with modest or frequent subsurface condition problems are distributed throughout the rest of the country. Most of the remaining questions in Part One address sub- surface investigation practices. As shown in Table 2, slightly more than half of responding agencies indicated that site characterization is often difficult. Among these agencies, most cited highly variable subsurface conditions rather than difficult-to-characterize materials as the source of site char- acterization difficulty. Responses from Table 2 are shown by agency in Figure 4. As noted for the frequency of subsurface conditions problems (see Figure 3), many of the agencies for which site charac- terization is generally not difficult are located in the Central Lowlands and Great Plains geologic regions of the Midwest. There is also a cluster of agencies in the Coastal Plain region of the South Atlantic states. Most of the responding agencies (40, 78%) have a state- specific manual and/or specifications that describe require- ments and practices for site characterization, and most of those agencies included an Internet link to access the manual. A list of the links is included in the short answer responses to Question 5 included with Appendix B3. All respondents indicated that it is important that sub- surface information be provided with project bid documents; a list of the required documents for each agency is also included with Appendix B3 (Question 6). The documents vary widely, with some agencies providing all available geotechnical information and reports and others limiting the information to boring logs only. Several agencies included Geotechnical Data Reports; however, none mentioned GBRs, which were discussed in chapter two. Several agencies also indicated that some geotechnical information is available for background information only.

14 FIGURE 2 Survey agencies. Response Number Percent My agency experiences relatively few design, construction, and performance problems resulting from subsurface conditions 17 33 My agency experiences a modest number of design, construction, and performance problems resulting from subsurface conditions 32 63 My agency experiences frequent design, construction, and performance problems resulting from subsurface conditions 2 4 51 responses. TABLE 1 AGENCY PROBLEMS FROM SUBSURFACE CONDITIONS FIGURE 3 Frequency of subsurface conditions problems from agency survey responses.

15 As shown in Table 3, minimum or suggested minimum subsurface investigation requirements for all projects are for- mally specified for most of the responding agencies (44, 86%). The other seven agencies do not formally specify minimum requirements. Among the 44 agencies that do specify require- ments, 28 specify minimum requirements that are equal to or generally consistent with those prescribed in AASHTO speci- fications and guidelines. Agency requirements for five agen- cies exceed AASHTO requirements. Three agencies indicated that their subsurface investigation requirements are materially different from the AASHTO specifications and one that its requirements are generally less stringent than the AASHTO specifications. The 44 agencies that specify minimum require- ments were also asked about the frequency of exceeding mini- mum requirements; with responses summarized in Figure 5. About one-half (23, 52%) responded that minimum require- ments were occasionally exceeded, one-third (15, 34%) rarely exceed minimum requirements, and the rest (6, 14%) reported it was common to exceed minimum requirements. Respondents were also asked about the use of geophysical methods (Figure 6) and how historic subsurface information was maintained (Figure 7). About one-half (27, 53%) of agen- cies reported occasionally using geophysical methods. Only five agencies (10%) reported they routinely use geophysical methods and only two (4%) never use geophysical methods. The rest (16, 31%) rarely use geophysical methods. Methods of maintaining historic subsurface information varied. Nearly one-third of responding agencies (16, 31%) maintain a geographic information system (GIS)-based data- base of subsurface information. Slightly more (20, 39%) main- tain a non-GIS-based electronic database, and the remaining agencies keep paper copies. Twelve of the 15 agencies that TABLE 2 DIFFICULTY OF SITE CHARACTERIZATION Response Number Percent Site characterization is generally not difficult 22 43 Site characterization is often difficult because of highly variable subsurface conditions 23 45 Site characterization is often difficult because select types of soil and rock are difficult to characterize 1 2 Site characterization is often difficult because of highly variable subsurface conditions and select types of soil and rock that are difficult to characterize 5 10 51 responses. FIGURE 4 Sources of site characterization difficulty from agency survey responses.

16 reported maintaining a paper database indicated that the his- toric records can be accessed when needed; the other three indicated accessing the historical records is difficult. NATURE OF CLAIMS, CHANGE ORDERS, AND COST OVERRUNS Parts Two and Three of the survey contained questions regard- ing the nature of claims, change orders, and cost overruns, including how problematic they are, how they’ve changed in the past 10 years, and how they are affected by project deliv- ery mechanism. Most of the questions were asked a first time with respect to all claims, change orders, and cost overruns before repeating the question a second time to inquire specif- ically about claims, change orders, and cost overruns that can be attributed to subsurface conditions or site characteriza- tion practices. Although Part Two asked primarily qualitative questions about claims, change orders, and cost overruns, Part Three specified quantitative information. The difficulty FIGURE 5 Frequency of exceeding minimum subsurface investigation requirements among the 44 agencies that specify requirements. FIGURE 6 Agency use of geophysics (51 responses). FIGURE 7 Agency maintenance of historic subsurface information (51 responses). Response Number Percent Minimum requirements are those prescribed in AASHTO specifications and guidelines 7 14 Minimum requirements are documented in agency-specific provisions, but are generally consistent with those prescribed in AASHTO specifications and guidelines1 28 55 Minimum requirements are documented in agency-specific provisions, but substantially exceed those prescribed in AASHTO specifications and guidelines1 5 10 Minimum requirements are documented in agency-specific provisions, but are materially different from those prescribed in AASHTO specifications and guidelines1 (e.g., involve different techniques and procedures than are addressed in AASHTO specifications) 3 6 General minimum requirements are documented in agency-specific publications, but are generally less stringent than those prescribed in AASHTO specifications and guidelines1 1 2 Minimum (or recommended minimum) requirements are not specified 7 14 51 responses. 1 “AASHTO specifications and guidelines” was the phrase used in the survey question. This could be interpreted as either the LRFD Bridge Design Specifications (2014), or the Manual on Subsurface Investigations (1988); the documents have essentially the same investigation requirements. TABLE 3 MINIMUM SUBSURFACE INVESTIGATION REQUIREMENTS

17 of collecting and interpreting quantitative information regard- ing claims, change orders, and cost overruns was discussed in chapter one; the results from qualitative and quantitative questions are presented respectively in the first two sections. A third section synthesizes the information from both sets of questions to address the extent of claims, change orders, and cost overruns that can be attributed to subsurface conditions. Qualitative Information Regarding Claims, Change Orders, and Cost Overruns Nearly half of respondents indicated that claims, change orders, and cost overruns from all sources are a recognized problem and priority concern for their agencies (Table 4). For the other responding agencies, responses skewed toward indi- cating that claims, change orders, and cost overruns are less problematic, although two agencies noted that they are one of their agency’s most significant problems. The sources of agency concerns are shown in Figure 8. Respondents were asked to select all sources of concern that apply from among the options listed in the figure. The concerns cited most fre- quently were agency budget and the time allocation of agency resources, although public perception was also a concern for more than one-third of respondents that indicated claims, change orders, and cost overruns are a recognized problem for their respective agencies. Agency level of concern specific to subsurface conditions claims, change orders, and cost overruns is shown in Table 5. Nearly 40% of responding agencies indicated that these con- cerns are not considered to be a significant problem for their agency. The other responding agencies were almost evenly Response Number Percent Claims, change orders, and cost overruns are not considered to be a significant problem 8 16 Claims, change orders, and cost overruns are recognized as a problem, but are not a priority concern 13 26 Claims, change orders, and cost overruns are recognized as a problem and are a priority concern 23 46 Claims, change orders, and cost overruns are recognized as one of our agency’s most significant problems 2 4 I don’t know 4 8 50 responses. TABLE 4 AGENCY LEVEL OF CONCERN REGARDING ALL CLAIMS, CHANGE ORDERS, AND COST OVERRUNS FIGURE 8 Agency sources of concern regarding all claims, change orders, and cost overruns. Respondents were allowed to select more than one response. In response to the previous question, 38 respondents indicated that claims, change orders, and cost overruns from all causes were a recognized problem for their respective agencies.

18 split between such claims being a recognized problem but not a priority concern and being a recognized problem and a priority concern. None of the respondents indicated claims, change orders, and cost overruns resulting from subsurface conditions are one of his or her agency’s most significant problems. The differences between responses regarding all claims, change orders, and cost overruns and those resulting from subsurface conditions are discussed further in Extent of Claims, Change Orders, and Cost Overruns Resulting from Subsurface Conditions later in this chapter. Respondents were also asked if the frequency and/or mag- nitude of claims, change orders, and cost overruns from all sources have changed since 2005. As shown in Table 6, the most common response (18 of 50 respondents) was that the respondent did not know. Among the remaining respondents, 15 indicated that the magnitude and/or frequency of claims, change orders, and cost overruns has remained about the same; the remaining 17 were about evenly divided between having increased and having decreased, with more respon- dents indicating the changes were slight rather than significant. Respondents were also asked about the change in magnitude and/or frequency of claims, change orders, and cost overruns resulting from subsurface conditions in the past ten years. The results are shown in Table 7. Interestingly, more respondents were confident enough to select an answer, with nearly half indicating the magnitude and/or frequency had stayed the same. Other respondents were again about evenly divided between having increased and having decreased, with more respondents indicating the changes were slight rather than significant. The differences between responses regarding all claims, change orders, and cost overruns and those resulting from subsurface conditions are discussed further in Extent of Claims, Change Orders, and Cost Overruns Resulting from Subsurface Conditions later in the chapter. Respondents were also asked if project delivery mecha- nism (design-bid-build, design-build, public-private partner- ship, construction manager/general contractor) is perceived to have a significant effect on the incidence or magnitude of claims, change orders, and/or cost overruns. The most common response, from 19 of 50 respondents (38%), was “I don’t know.” Among the other respondents, 17 (34%) reported that there is no effect and 14 (28%) that there is. These 14 respondents were Response Number Percent Claims, change orders, and cost overruns resulting from subsurface conditions are not considered to be a significant problem 18 38 Claims, change orders, and cost overruns resulting from subsurface conditions are recognized as a problem, but are not a priority concern 15 31 Claims, change orders, and cost overruns resulting from subsurface conditions are recognized as a problem and are a priority concern 13 27 Claims, change orders, and cost overruns resulting from subsurface conditions are recognized as one of our agency’s most significant problems 0 0 I don’t know 2 4 48 responses. TABLE 5 AGENCY LEVEL OF CONCERN REGARDING CLAIMS, CHANGE ORDERS, AND COST OVERRUNS RESULTING FROM SUBSURFACE CONDITIONS Response Number Percent The magnitude and/or frequency has decreased significantly 4 8 The magnitude and/or frequency has decreased slightly 5 10 The magnitude and/or frequency has remained about the same 15 30 The magnitude and/or frequency has increased slightly 5 10 The magnitude and/or frequency has increased significantly 3 6 I don’t know 18 36 50 responses. TABLE 6 CHANGES IN CLAIMS, CHANGE ORDERS, AND COST OVERRUNS FROM ALL SOURCES IN THE PAST TEN YEARS Response Number Percent The magnitude and/or frequency has decreased significantly 2 4 The magnitude and/or frequency has decreased slightly 6 12 The magnitude and/or frequency has remained about the same 22 45 The magnitude and/or frequency has increased slightly 6 12 The magnitude and/or frequency has increased significantly 1 2 I don’t know 12 24 49 responses. TABLE 7 CHANGE IN CLAIMS, CHANGE ORDERS, AND COST OVERRUNS RESULTING FROM SUBSURFACE CONDITIONS IN THE PAST TEN YEARS

19 asked to explain the effect. The responses are included with the short answers in Appendix B. In their answers, respondents primarily referred to design-build versus design-bid-build, with five mentioning there are fewer claims, change orders, and cost overruns for design-build than for design-bid-build and four stating there are more. The other five respondents did not indicate whether the effect is negative or positive. Many respondents cited a change in risk allocation in their responses. Respondents were also asked about the effect of project delivery mechanism for claims, change orders, and cost over- runs related to subsurface conditions. The most common answer was that there is no effect, cited by 22 of 49 respon- dents (45%). Sixteen respondents (33%) selected “I don’t know” and 11 (22%) indicated there is an effect. These 11 were asked to explain this effect. The responses are included with the short answers in Appendix B. As for the question regarding all claims, change orders, and cost overruns, the respondents primarily referred to design-build versus design- bid-build, and the responses were split regarding whether the effect is negative or positive, with four respondents indicat- ing there are more claims, change orders, and cost overruns for design-build than design-bid-build, and four that there are fewer. Three of the responses did not mention whether the effect is negative or positive. Many of the responses cited the same explanations given for all claims, although two respondents noted that there is less geotechnical information collected for design-build projects. Quantitative Information Regarding Claims, Change Orders, and Cost Overruns In Part Three of the survey, respondents were asked for spe- cific quantitative information regarding the claims, change orders, and cost overruns that had occurred for their agencies during the past five years. As discussed in chapter one, such quantitative information is difficult to collect. The quantita- tive results presented in this section are therefore limited to data reported by 11 agencies. The discussion in chapter one also addressed the difficulties of interpreting such quan- titative data. The difficulties largely relate to distinguish- ing those resulting from geotechnical investigation issues from those resulting from design, communication, or other issues. All values cited in the body of this synthesis report have been averaged over the 5-year period (2009–2013) to report values in annual terms (e.g., if an agency reported that its change orders between 2009 and 2013 totaled $5 million, the average annual change order total would be $1 million). Information requested included the total number and total cost of each of the three categories (claims, change orders, and cost overruns), as well as the total contract costs for all projects on which the claims, change orders, and cost over- runs were recorded, respectively. Eleven respondents completed Part Three of the survey. Summary statistics for the results are presented in Tables 8–10. TABLE 8 SUMMARY STATISTICS FOR CLAIMS, CHANGE ORDERS, AND COST OVERRUNS FROM ALL SOURCES Total Number Total Cost Statistic Claims Change Order Cost Overruns Claims Change Order Cost Overruns Number of Responses 8 9 6 8 9 7 Minimum 0.8 150 10 $0.02* $5.1* $2.6* Maximum 75 6,355 209 $12.5* $82.5* $70.4* Average 18 1,688 107 $3.1* $37.8* $29.8* Standard Deviation 30 1,928 84 $4.7* $27* $23.5* Values are average annual values for data from 2009 to 2013. *In millions. TABLE 9 SUMMARY STATISTICS FOR CLAIMS, CHANGE ORDERS, AND COST OVERRUNS ATTRIBUTED TO SUBSURFACE CONDITIONS Total Number Total Cost Statistic Claims Change Order Cost Overruns Claims Change Order Cost Overruns Number of Responses 4 5 2 4 5 2 Minimum 0 6 3 $0 $1.0* $0.3* Maximum 7.8 125 21 $865,000 $10.3* $1.0* Average 2 46 12 $240,000 $3.4* $0.7* Standard Deviation 3.9 49 13 $419,000 $3.9* $0.5* Values are average annual values for data from 2009 to 2013. *In millions.

20 TABLE 10 AVERAGE COST PER OCCURRENCE OF CLAIMS, CHANGE ORDERS, AND OVERRUNS All Causes Subsurface Conditions Only Claims $169,000 $120,000 Change Orders $22,000 $74,000 Overruns $279,000 $55,000 TABLE 11 AVERAGE ANNUAL BUDGET FOR NEW CONSTRUCTION FOR ALL AGENCIES PROVIDING ANY RESPONSES TO PART 3 Agency A ve ra ge A nn ua l Co ns tru ct io n B ud ge t Arkansas $526* Eastern Federal Lands $132* Florida — Georgia $850* Indiana $1,196* Maryland $493* Massachusetts $1,000* Missouri $960* New Hampshire $277* Oregon $928* South Dakota $325* Data are from 2009 to 2013. — = Data not provided in agency response. *In millions. TABLE 12 AVERAGE ANNUAL DATA REGARDING CLAIMS FROM ALL SOURCES FOR DATA FROM 2009 TO 2013 Agency N um be r A nn ua l C os t A ve ra ge C os t Co st as S ha re o f A ge nc y Co ns tru ct io n Bu dg et To ta l C on tra ct C os ts fo r Pr oje cts w ith C lai ms A ve ra ge C la im C os t a s Sh ar e of P ro jec t B ud ge t Arkansas 2 $0.3* $140,000 0.05% $11* 2.5% Eastern Federal Lands 2 $0.3* $142,000 0.24% $8* 3.8% Georgia 6 $3* $500,000 0.35% $50* 6.0% Massachusetts 75 $12.5* $166,000 1.25% — — Missouri 1 $0.2* $240,000 0.02% $2* 10.0% New Hampshire 1 $0.02* $19,000 0.01% $3* 0.6% Oregon 58 $8.2* $143,000 0.89% $145* 5.7% South Dakota 2 $0.3* $183,000 0.10% — — *In millions. More detailed summary results are presented in Tables 11–14, as well as in Tables 15–17 for claims, change orders and cost overruns that can be attributed to subsurface conditions or site characterization practices. In addition, all data reported by the respondents are included in Appendix B. Four of the 11 responses did not include information specific to sub- surface conditions, and several did not include information for all three categories of claims, change orders, and cost over- runs. These missing data are indicated in the tables as empty cells. The tables also present the average cost of an individ- ual occurrence for each category (e.g., the average cost of a change order) and percentages representing the cost of each category relative to the total agency budget for new construc- tion and the cost of each category relative to total project costs. Several noteworthy observations related to the magnitude and variability of the values presented in Tables 8–17 are listed here: • Considering all causes, change orders occur most fre- quently, with an average annual rate of occurrence of 1,638, more than ten times either cost overruns (126) or claims (88). • Total amounts spent on all change orders and cost over- runs are considerably greater than that spent on all claims. Change orders and cost overruns typically amount to 3% to 5% each of an agency’s total construction budget and as much as 18%. • As observed for all sources, subsurface conditions change orders occur most frequently, with an average annual rate of occurrence (46) significantly more than either cost overruns (12) or claims (2). • The total amounts spent on subsurface conditions change orders is considerably greater than that spent on claims or cost overruns attributed to subsurface conditions. Reported spending on change orders resulting from sub- surface conditions approached 1% of the total agency budget for new construction. • The average value of each individual claim, change order, and cost overrun is relatively consistent from

21 TABLE 13 AVERAGE ANNUAL DATA REGARDING CHANGE ORDERS FROM ALL SOURCES FOR DATA FROM 2009 TO 2013 Agency N um be r A nn ua l C os t A ve ra ge C os t Co st as S ha re o f A ge nc y Co ns tru ct io n Bu dg et To ta l C on tra ct C os ts fo r Pr oje cts w ith C ha ng e O rd er s A ve ra ge C ha ng e O rd er Co st as S ha re o f P ro jec t B ud ge t Arkansas 1,028 $26* $25,000 4.9% $797* 3.3% Eastern Federal Lands 150 $23* $150,000 17.4% $225* 10.2% Florida 6,355 $83* $13,000 — $2,694* 3.1% Georgia 420 $50* $119,000 5.9% $1,100* 4.6% Indiana 2,699 $61* $23,000 5.1% $1,196* 5.1% Maryland — $25* — 5.2% — — Missouri 1,916 $5.1* $3,000 0.53% $1,200* 0.42% New Hampshire 465 — — — — — Oregon 1,426 $61* $43,000 6.6% $928* 6.6% South Dakota 730 $6.9* $9,000 2.1% — — — = Data were not provided in agency response. *In millions. TABLE 14 AVERAGE ANNUAL DATA REGARDING COST OVERRUNS FROM ALL SOURCES FOR DATA FROM 2009 TO 2013 Agency N um be r A nn ua l C os t A ve ra ge C os t Co st as S ha re o f A ge nc y Co ns tru ct io n Bu dg et To ta l C on tra ct C os ts fo r Pr oje cts w ith C ost O ve rru ns A ve ra ge O ve rru n Co st as Sh ar e o f P ro jec t B ud ge t Arkansas 134 $16.0* $119,000 3.0% $372* 4.3% Eastern Federal Lands 24 $23* $930,000 17.1% $122* 18.5% Florida 209 $51* $241,000 — $1,550* 3.3% Georgia — $14* — 1.7% $1,080* 1.3% Missouri 189 $33* $174,000 3.4% $653* 5.0% New Hampshire 10 $2.6* $264,000 0.9% $56* 4.6% Oregon 74 $70* $951,000 7.6% $431* 16.4% — = Data were not provided in agency response. *In millions. TABLE 15 AVERAGE ANNUAL DATA REGARDING CLAIMS RESULTING FROM SUBSURFACE CONDITIONS FOR DATA FROM 2009 TO 2013 Agency N um be r A nn ua l C os t A ve ra ge C os t Co st as S ha re o f A ge nc y Co ns tru ct io n B ud ge t To ta l C on tra ct C os ts fo r P ro jec ts wi th Cl ai m s A ve ra ge C la im C os t as S ha re o f P ro jec t B ud ge t Eastern Federal Lands 0.2 $93,000 $467,000 0.07% $1.59* 5.9% Georgia 0 N/A N/A N/A N/A N/A Massachusetts 7.8 $865,000 $111,000 0.09% — — South Dakota 0 N/A N/A N/A N/A N/A N/A = Not applicable since zero claims were filed. — = Data were not provided in agency response. *In millions.

22 one agency to the next, when considering all causes as well as only those attributed to subsurface conditions. The average values are reported in Table 10. The mag- nitudes of these values suggest that the effect of claims, change orders, and cost overruns on individual project budgets could be substantial. When all causes are con- sidered, overrun costs are most significant and change orders are considerably less than both claims and over- runs. When only subsurface conditions are considered, there is less difference in average cost per occurrence. • Variation in the numbers and costs of claims, change orders, and cost overruns reported to be considerable. This is understandable, because agencies have different program sizes, different practices, and different sub- surface conditions, among other factors. • Variation in the relative proportion of claims, change orders, and cost overruns may also suggest agencies have different procedures for reconciling disputes. EXTENT OF CLAIMS, CHANGE ORDERS, AND COST OVERRUNS RESULTING FROM SUBSURFACE CONDITIONS Uncertainty regarding the significance of subsurface condi- tions, in terms of overall cost and as a share of all claims, change orders, and cost overruns, is substantial. The infor- mation gathered from Parts Two and Three of the survey and summarized in the previous sections can be used to reduce some of the uncertainty, which is a principle objective of this Synthesis project. Comparing the responses to the qualitative survey questions about the level of agency concerns is one method for evaluating the extent of claims, change orders, and cost overruns resulting from subsurface conditions. The responses to those questions were presented earlier and are summarized in Figure 9. On one hand, 12 of the 25 agencies that consider claims, change orders and cost overruns from all causes a priority concern or a most significant problem do not have the same degree of concern for those attributed to subsurface conditions; on the other hand, 13 agencies do consider subsurface conditions a priority con- cern, and 13 of 48 respondents (27%) acknowledge widespread concern. Concern for claims, change orders, and cost overruns resulting from subsurface conditions could never exceed con- cern for all causes: among the 25 agencies that consider claims, change orders, and cost overruns a priority or greater concern, 13 consider subsurface conditions to be an equal concern. That more than half of the agencies are concerned in particular about subsurface conditions is noteworthy. The responses to qualitative survey questions about the ten-year change in claims, change orders, and cost overruns TABLE 16 AVERAGE ANNUAL DATA REGARDING CHANGE ORDERS RESULTING FROM SUBSURFACE CONDITIONS FOR DATA FROM 2009 TO 2013 Agency N um be r A nn ua l C os t A ve ra ge C os t Co st as S ha re o f A ge nc y Co ns tru ct io n B ud ge t To ta l C on tra ct C os ts fo r P ro jec ts wi th Ch an ge O rd er s A ve ra ge C ha ng e O rd er C os t a s S ha re o f P ro jec t B ud ge t Eastern Federal Lands 5.6 $1.0* $172,000 0.73% $14.5* 6.6% Georgia 42 $2.0* $48,000 0.24% $1,080* 0.19% Indiana 125 $10* $83,000 0.86% $585* 1.8% Maryland 6.8 $1.5* $223,000 0.31% — — Oregon 49.2 $2.2* $45,000 0.24% $299* 0.73% — = Data were not provided in agency response. *In millions. TABLE 17 AVERAGE ANNUAL DATA REGARDING COST OVERRUNS RESULTING FROM SUBSURFACE CONDITIONS FOR DATA FROM 2009 TO 2013 Agency N um be r A nn ua l C os t A ve ra ge C os t Co st as S ha re o f A ge nc y Co ns tru ct io n Bu dg et To ta l C on tra ct C os ts fo r Pr oje cts w ith C ost O ve rru ns A ve ra ge O ve rru n Co st as S ha re o f P ro jec t B ud ge t Eastern Federal Lands 3 $0.3* $104,000 0.24% $9.61* 3.2% Georgia 21 $1.0* $10,000 0.12% $120* 0.83% *In millions.

23 from all causes and from subsurface conditions are plotted side by side in Figure 10. More agencies have seen no sig- nificant change than have observed a change, and the number of agencies that have observed an increase is approximately equal to the number that has observed a decrease. This sug- gests that, on average, claims, change orders, and cost over- runs have neither increased nor decreased during the last ten years. There appears to be more uncertainty regarding the ten-year change in all sources than the change in those attrib- uted to subsurface conditions, with 18 respondents selecting “I don’t know” for all sources and 12 “I don’t know” for subsurface conditions. That claims, change orders, and cost overruns, including those resulting from subsurface conditions, have not, on aver- age, changed considerably in ten years could indicate that their significance has remained steady; however, there are other possible explanations. Changes in construction practices and technology have likely affected the nature of claims, change orders, and cost overruns, reducing some causes while intro- ducing new ones. Changes in construction program size could also affect their significance. In Part Two of the survey, respondents were also asked to estimate the proportion of claims, change orders, and cost FIGURE 9 Agency concern regarding claims, change orders, and cost overruns. FIGURE 10 Perceived ten-year change in claims, change orders, and cost overruns.

24 overruns that can be attributed to subsurface conditions or site characterization practices. The responses are presented in Table 18. Nearly two-thirds of respondents (30 of 47) estimated that less than 20% can be attributed to subsurface conditions, with 14 agencies estimating the percentage to be between 20% and 40% and three estimating between 40% and 60%. Responses from Table 18 are shown by agency in the map of Figure 11. Geographic trends are less apparent than those observed based on the distribution of agencies experiencing subsurface conditions (Figure 3) and site characterization dif- ficulties (Figure 4). Three of the 14 agencies estimating 20% to 40% resulting from subsurface conditions are in Southeast- ern states, but otherwise the geographic distribution of the estimates appears to be random. The estimates from Part Two can be evaluated by consid- ering the quantitative information regarding claims, change orders, and cost overruns presented in the previous section (Tables 12–17). Figure 12a presents the number of claims experienced annually by the four agencies that provided data for both all claims and for claims that could be attributed to subsurface conditions or site characterization practices. Fig- ures 12b and c present similar information, but for change orders and cost overruns, respectively. The plots of Figure 13 are similar to those of Figure 12; however, for costs (rather than numbers) of claims (Figure 13a), change orders (Fig- ure 13b), and overruns (Figure 13c). The results shown in Figures 12 and 13 include percent- age values for each agency corresponding to the proportion of claims, change orders, and cost overruns that are attributed to subsurface conditions. These values are repeated in Table 19 by number on the left side and by cost the right side. Aver- age values are also shown. The first row of average values is the sum of all reported subsurface conditions claims, change orders, and cost overruns, respectively, out of all claims, change orders, and cost overruns. The second row of average values is similar to the first, but is irrespective of category; all claims, change orders, and cost overruns are considered together. Both rows of averages are averages of all data rather than averages of agency values. The second row of averages shown in Table 19 is especially important. By number, 5% of all reported claims, change orders, and cost overruns were attributed to subsurface con- ditions; by cost, 7% were attributed to subsurface conditions. The calculated shares support the majority response shown in Table 18 that less than 20% of claims, change orders, and overruns are attributed to subsurface conditions. There is variation in the potential interpretations of the quantitative results from Part Three of the survey. On one hand, subsurface conditions account for a relatively small percentage Response Number Percent Less than 20% 30 64 20%–40% 14 30 40%–60% 3 6 60%–80% 0 0 Greater than 80% 0 0 TABLE 18 AGENCY ESTIMATE OF PROPORTION OF CLAIMS, CHANGE ORDERS, AND COST OVERRUNS CAUSED BY SUBSURFACE CONDITIONS FIGURE 11 Agency estimate of proportion of claims, change orders, and cost overruns caused by subsurface conditions.

25 of claims, change orders, and cost overruns; 5% by number and 7% by cost. On the other hand, several measures of the total cost of subsurface conditions indicate that their cost is substan- tial. This annual cost of subsurface conditions change orders alone was commonly in the millions of dollars and as much as $10 million for the reporting agencies, representing a consider- able portion of individual project budgets, especially relative to project contingency funds, which are often limited. The costs also consume a significant portion of agencies’ total budgets for new construction. The two agencies that reported total costs for all categories, claims, change orders, and cost overruns attrib- uted to subsurface conditions summed to 0.35% and 1.0% of the total agency budget for new construction. PERCEIVED RELATIONSHIP BETWEEN SUBSURFACE INVESTIGATION PRACTICES AND CLAIMS, CHANGE ORDERS, AND COST OVERRUNS The survey results also shed light on the effect of subsurface investigation practices on reducing subsurface conditions claims, change orders, and cost overruns, which is another principle objective of this Synthesis project. The topic is examined in this section by considering information pro- vided in Part One regarding the effect of any recent changes to agency subsurface investigation practices and the effect of agency practices that incentivize subsurface investiga- tion. The effect of subsurface investigation practices is addressed further in chapter four, which examines in detail specific practices from five agencies selected for further examination. In Part One of the survey, 18 respondents (35%) indi- cated that their agencies had implemented specific changes to site characterization practices during the last five years. The respondents’ summaries of the changes are included with Appendix B. Three common revisions were each cited by at least four of the 18 agencies: increased use of CPT, modifications associated with agency design policies transitioning from allowable stress design to Load Factor and Resistance Design (LFRD), and increased frequency of explorations. These respondents were asked about the effect of such changes on the number of claims, change orders, and cost overruns, and the results are shown in FIGURE 12 Annual number of (a) claims, (b) change orders, and (c) cost overruns. a b c

26 a b c FIGURE 13 Annual cost of (a) claims, (b) change orders, and (c) cost overruns. Percentage By Number Percentage By Cost Claims Change Orders Cost Overruns Claims Change Orders Cost Overruns Eastern Federal Lands 10% 4% 12% 30% 4% 1% Georgia 0% 10% — 0% 4% 7% Indiana — 5% — — 17% — Maryland — — — — 6% — Massachusetts 10% — — 7% — — Oregon — 3% — — 4% — South Dakota 0% — — 0% — — Average, All Responses 9.4% 4.7% 12% 5.9% 7.8% 3.6% Average, All Responses and All Categories 4.9% 7.2% — = Data were not provided in agency response. TABLE 19 PERCENTAGES OF CLAIMS, CHANGE ORDERS, AND COST OVERRUNS ATTRIBUTED TO SUBSURFACE CONDITIONS

27 Table 20. Nearly one-quarter of respondents whose agen- cies had an updated site characterization practice indicated a marginal decrease in the occurrence of claims, change orders, and overruns, and two agencies (11%) indicated a noticeable decrease. These two agencies, Florida and Indi- ana, are discussed in case examples in chapter four. Eight respondents (16%) noted that their agency’s design code contains provisions that incentivize specific site char- acterization activities and/or performing site characteriza- tion in excess of minimum requirements. The respondents were asked to summarize the provisions and their descrip- tions are included in Appendix B. The respondents’ per- ception of the success of these activities is summarized in Table 21. None of the eight respondents indicated that the provisions had resulted in fewer claims, change orders, or overruns; however, cite design efficiencies resulting from the use of greater resistance factors (or lower factors of safety). SUMMARY OF SIGNIFICANT FINDINGS • Nearly 70% of responding agencies specify minimum subsurface investigation requirements that are equal to or generally consistent with requirements prescribed in the AASHTO LRFD Bridge Design Specifications (2014) and the AASHTO Subsurface Investigation Manual (1988). Among the other agencies, the most common response was minimum requirements that substantially exceed those prescribed in the AASHTO Specifications. • Approximately half of the responding agencies reported that site characterization is difficult because of highly variable subsurface conditions, and approximately half reported that site characterization is not difficult. A small percentage of agencies also indicated that site characterization is difficult because select types of soil and rock are difficult to characterize. • In the Midwest, some agencies in the Central Lowlands and Great Plains geologic regions reported experienc- ing fewer problems relating to subsurface conditions and less difficulty with site characterization than most other agencies. In the South Atlantic, some agencies in the Coastal Plain geologic region reported less difficulty with site characterization than most other agencies. • Three-quarters of responding agencies recognize claims, change orders, and cost overruns from all sources as a significant problem. Slightly more than half recognize claims, change orders, and cost overruns attributed to subsurface conditions as a significant problem. • Most responding agencies perceive that the magnitude and frequency of claims, change orders, and cost over- runs has remained steady over the last decade, although some have noticed a decrease and others an increase in the magnitude and frequency. TABLE 20 PERCEIVED IMPACT OF REVISIONS TO SITE CHARACTERIZATION PRACTICES ON CLAIMS, CHANGE ORDERS, AND OVERRUNS Response Number Percent Changes have led to a noticeable increase in the occurrence of claims, change orders, and overruns 0 0 Changes have led to a marginal increase in the occurrence of claims, change orders, and overruns 0 0 Changes have not noticeably affected the occurrence of claims, change orders, and overruns 7 39 Changes have led to a marginal decrease in the occurrence of claims, change orders, and overruns 4 22 Changes have led to a noticeable decrease in the occurrence of claims, change orders, and overruns 2 11 Insufficient experience to respond 5 28 TABLE 21 PERCEIVED SUCCESS OF SITE CHARACTERIZATION INCENTIVES TO REDUCING CLAIMS, CHANGE ORDERS, AND OVERRUNS Response Number Percent Practice has not noticeably affected the occurrence of claims, change orders, and overruns 4 50 Practice has produced marginal reduction in the occurrence of claims, change orders, and overruns 0 0 Practice has produced substantial reduction in the occurrence of claims, change orders, and overruns 0 0 Insufficient experience to respond 2 25 I don’t know 2 25

28 • Change orders attributed to subsurface conditions are significantly more frequent than claims or cost overruns. The same trend was observed among claims, change orders, and cost overruns resulting from all causes. There was a less significant difference among the aver- age cost per individual occurrence of subsurface condi- tions claims, change orders, and cost overruns, which is contrary to the trend observed considering all causes, for which change orders were considerably less costly on average than claims and cost overruns. • By number, 5% of all claims, change orders, and cost overruns reported in the survey were attributed to sub- surface conditions, and by cost, 7% were attributed to subsurface conditions. • Cumulative costs for claims, change orders, and cost overruns attributed to subsurface conditions represented up to 1% of total agency construction budgets. • Individual, project-level data were not collected; how- ever, total annual agency costs of subsurface con- ditions change orders, for example, represented as much as 7% of the total agency spending on all proj- ects associated with the change orders. The impact on some individual project budgets is likely much greater than 7%.