Geotechnical Site Investigations for Underground Projects: Volume 1 (1984)

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. Interpretation of Case Histories The original data for the 87 projects discussed herein were obtained through an extensive procedure (see Appendix C) involving extraction of detailed information from documents submitted by owners and personal in- terviews with staff representing the owner, contractor, and engineers for each project. The procedure was extremely complicated, involving accurate recording of both qualitative and quantitative data. For exam- ple, qualitative data included statements regarding problems related to ground conditions encountered during construction, comments on the ef- fectiveness of the site investigation program, and opinions concerning disputes. The quantitative data covered a wide range of items, such as project costs and schedule, tunnel specifications, geologic criteria, types and number of exploration techniques, and construction methods and progress. Subsequently, some basic calculations using these quantita- tive data were made, for example to derive the face area, volume in cu- bic yards of excavation, borehole spacing, advance rate per shift, etc. Occasionally there was some overlapJ sometimes the actual excavated vol- ume in cubic yards had also been obtained from documents supplied by the owner. The original data from documents and interviews were recorded on a 15-page data form, with the interviewers often adding several pages of explanatory information. These data were then combined with the basic calculations. Thus, there was a large and complex body of qualitative and quantitative information to be examined. CBAR'1'BD AND PLOTTED DATA Su.aary Matrixes An array of geotechnical problems that occurred during construction of the 84 mined tunnels and 3 deep shafts can be seen at a glance in the summary matrix presented separately as Plate 1. The matrix shows the 87 study projects plotted against the abbreviated •problems encountered• list from the project abstracts. This list allowed for consideration of 31 separate items grouped into 7 categories: unstable ground, ground- water inflow, hazardous environmental factors, mechanical problems (rock and TBMs), soft-ground methods, compressed air, and other. 91 

Through the use of symbols, the matrix indicates which conditions developed into problems and which of the problems were serious enough to cause claims. The matrix makes it clear that most projects encounter not just one, but several construction problems. What may not be ap- parent is that many conditions interact or affect each other, so that some judgment was required in deciding on the primary culprits in an abbreviated list of problem descriptions. A second matrix was prepared to chart selected, original numerical data for each of the 87 projects in combination with basic calculated data for each. A total of 57 different items were displayed in the summary matrix, as shown in Table 7.1. TABLE 7.1 Contents of Data Matrix Original Data Calculated Data Name of project Months to build Purpose of tunnel Factor to escalate costs Number of bidders Cost, bid as t of estimate Start and finish dates Cost, total as t of estimate Cost, engineer's estimate Face area Cost, bid Tunnel volume Cost, total to build Cost, $/cu yd Cost, exploration Cost; $/lin ft Number of tubes Exploration, t of tunnel cost Length of tunnel Exploration, $/cu yd Shape of tunnel Boreholes, average depth Tunnel volume Boreholes, lin ft/route ft Type of ground Boreholes per 1,000 route ft Geology (simplified) Boreholes, spacing overburden, max and min Boreholes, $/lin ft water head, max and min overall advance per day Water inflow, max and min Advance per 8-hr shift Boreholes, number Labor, total man hours Boreholes, lin ft Labor, man hours/day Borehole depth, max and min Excavation, man hours/cu yd Boreholes, distance from centerline Excavation, cu yd/hr Compressive strength tests, number Claims made, $/cu yd Compressive strength, max and min Claims settled, $/cu yd Construction equipment/method Claims, $ settled as t made Primary support Advance per day, max and average Days worked Shifts worked Crew size Problems, construction Liquidated damages in specifications Claims made, $ Claims settled, $ The information summarized in the data matrix served for initial re- view of comprehensive results, following which the content of the matrix was expanded by additional basic calculations. This revised data summary 92 

formed the basis for plotting graphically and for arithmetic tabulation. Among the tabulations were such items as total claims made and settled, cost in dollars per cubic yard for different methods of excavation, boreholes in linear feet for tunnels in mountainous areas, etc. A modified version of the data summary matrix is presented sepa- rately as Plate 2. It displays 20 of the 57 items of original and calculated information contained in the complete summary. Plots Generated During several subcommittee meetings, the complete data summary was re- viewed in conjunction with the problem summary in order to select items that appeared to be most promising for study. To accommodate the variety of data selected for examination, the ability to sort and plot graphi- cally at various scales was essential. A specialized computer program was written by personnel of Tudor Engineering Company to select and plot TBM, drill-and-blast, soft-ground and compressed-air tunnels built for rapid transit, railroads or water conveyance, or underground . subway stations or hydroelectirc powerhouses. Table 7.2 lists the plots gen- erated to sort the data according to various combinations of parame- ters. Another computer program without plotting capability, prepared at Virginia Polytechnic Institute to manage the abstract form of the case histories, was also used to search and review the data on a general basis. TABLE 7.2 Data Plots Generated for Correlation Plot X Axis Y Axis Method 1 Boreholes, LF/RF Cost, 1982 $/cu yd All 2 Exploration, as t cost Total cost, as ' All eng. est. 3 Exploration, $/cu yd Cost, 1982 $/cu yd All 4 Boreholes, number Cost, 1982 $/cu yd All 5 Avg. Advance, LF/day Cost, 1982 $/cu yd All 6 Claims made, $ Total cost, $ All 7 Boreholes, LF/RF Total cost, as t All eng. est. 8 Water inflow, max gpm Total cost, as \ All eng. est. 9 Bidders, number Total cost, as ' All eng. est. 10 Boreholes/1,000 RF Cost, 1982 $/cu yd All 11 Face area Cost, 1982 $/cu yd All 12 water inflow, max gpm Cost, 1982 $/cu yd All 13 Avg. Advance, LF/day Cost, 1982 $/cu yd All 14 Boreholes, LF/RF Exploration, as \ cost All 15 Length, LF Cost, 1982 $/cu yd All 16 Advance rate, LF/day Cost, 1982 $/cu yd TBM, D&B 17 Advance rate, LF/day Cost, 1982 $/cu yd D&B 18 Advance rate, LF/day Cost, 1982 $/cu yd TBM 19 Advance rate, LF/day Cost, 1982 $/cu yd Soft ground 93 

TABLE 7.2 Data Plots (continued) ~ X Axis Y Axis Method 20 Boreholes, number Cost, as t eng. eat. TBM 21 Length, LF Cost, as t eng. est. TBM 22 Avg. Advance, LP/day Cost, as t eng. est. TBM 23 Exploration, as ' coat Cost, 1982 $/cu yd TBM 24 Boreholea/1,000 RP Total cost, as t TBM eng. est. 25 Boreholes, number Total cost, as t D&B eng. est. 26 Length, LP Cost, as t eng. est. D&B 27 Avg. Advance, LP/day Cost, as t eng. est. D&B 28 Exploration, as t cost Cost, 1982 $/cu yd D&B 29 Boreholes/1,000 RF Cost, as t eng. est. D&B 30 Boreholes, number Total cost, as t All eng. est. 31 Length, LP Cost, as ' eng. est •• All 32 Avg. Advance, LF/day Cost, as \ eng. est. All 33 Exploration, as t cost Cost, 1982 $/cu yd All 34 Length, LP Cost, as ' eng. est. All 35 Boreholes/1,000 RP Total cost, as ' All eng. est. 36 Exploration, as t cost Claims paid, as ' All total cost 37 Boreholes, LP/RF Claims paid, as t All total cost 38 Boreholes/1,000 RF Claims paid, as t All total cost 39 Boreholes, number Claims paid, as t All total cost 40 Exploration, as t cost Claims made, as t coat All 41 Boreholes, LP/RF Claims made, as ' coat All 42 Boreholes/1,000 RF Claims made, as \ cost All 43 Boreholes, number Claims made, as ' coat All 44 Length, LP Cost, 1982 $/cu yd TBM 45 Boreholes, LF/RF Cost, 1982 $/cu yd TBM (water) 46 Total cost Cost, as \ eng. est. All 47 Boreholes, LF/RF Avg. Advance/8 hr All 48 OVerburden, max Avg. Advance/8 hr All 49 Avg. Advance/8 hr Length D&B so Avg. Advance/8 hr Cost, 1982 $/cu yd D&B 51 Avg. Advance/8 hr Face area D&B 52 Excavation, cu yd/hr Length D&B 53 Excavation, cu yd/hr Cost, 1982 $/cu yd D&B 54 Excavation, cu yd/hr Face area D&B 55 Avg. Advance/8 hr Length TBM 56 Avg. Advance/8 hr Cost, 1982 $/cu yd TBM 57 Avg. Advance/8 hr Face area TBM 58 Excavation, cu yd/hr Length TBM 59 Excavation, cu yd/hr Face area TBM 60 Excavation, cu yd/hr Cost, 1982 $/cu yd TBM 94 

TABLE 7.2 Data Plots (continued) Plot X Axis Y Axis Method 61 Boreholes, LF/RF Bid, as ' eng. est. All 62 Boreholes LF/RF Bid, as ' total cost All 63 Boreholes, LF/RF Claims made, as All 64 Boreholes, LF/RF eng. est. Claims made, as ' bid All 65 66 Boreholes, LF/RF Exploration, as ' Total cost, as ' bid Bid, as ' total cost All All 67 eng. est. Exploration, as ' Total cost, as ' All eng. est. ' eng. est. 68 Exploration, as Claims made, as All 69 eng. est. Exploration, as ' eng. est. Claims made, as ' bid All 70 eng. est. Exploration, as ' Total cost, as ' ' All 71 total cost Boreholes, LF/RF ' eng. est. Total project cost, as All ' eng. est. 72 Boreholes, LF/RF Total project cost, as All ' bid •cost• refers to mined tunnel (or shaft) construction only, excluding claims and modifications awarded. •Total cost• is synonymous with •as completed cost, • which includes any claims and modifications awarded. •Project cost• refers to the total contract, of which the mined tunnel is a part. Review of the plots led to a determination that many of the combina- tions of parameters reflected a lack of significant or meaningful corre- lation. Moreover, in cases where the parameters had been further sorted for plotting according to construction method, sampling limitations pro- duced results that were deemed generally inadequate for correlation pur- poses. The ability to distinguish among the types of projects proved interesting for discussion of the plots but provided no conclusive re- sults. Variation in sample size was a continuing concern because of its po- tential for limiting or negating the utility of the plots. As noted above, the sorting technique was a factor that ultimately yielded inade- quate samples for more than 30 percent of the plots. However, the plots generated to examine all projects were also subject to some reduction in sample size arising from availability of data for parameters. The ef- fects were most apparent for parameters based on water inflow and on excavation and advance rates combined with work force units (number of length of shifts and crew size). Results for other parameters, such as exploration costs, were monitored for possible sampling influence. From the standpoint of availability of samples, the parameters considered most reliable for correlation purposes included data relating to tunnel length, face area, overburden, cubic yards of excavation, number of bid- ders, boreholes, engineer's estimate, bid estimate, total cost, and claims made and awarded. 95 

The subcommittee critically reviewed the results derived from the mass of qualitative and quantitative information gathered from the 87 case histories, and selected the more distinct of the pertinent results for presentation. The discussion that follows is confined to matters that bear on the nature of the relationship between geology and con- struction and the significance of the geotechnical site investigation. IR'RRPRBTATIOIJ OP RESULTS Data derived from the 84 mined tunnels included in the 87 study projects shown in Plate 1 is tabulated in Table 7.3. Overall, unstable ground is the most prevalent problem encountered during construction, with blocky/slabby and running ground cited most often as specific conditions (38 percent and 27 percent, respectively) for all the projects. Ground- water inflow is cited as a problem in 33 percent of the projects. TABLE 7.3 Problems and Claims* Reported for Mined Tunnels Blocky/slabby rock, overbreak, cave-ins ,, Problems of tunnelsl 38 ,, 16 ClaiiiS of tunnelsl Running ground 27 9 Flowing ground 5 4 Squeezing ground 19 8 Spalling, rock bursts 6 4 Groundwater inflow 33 6 Noxious fluids 6 4 Methane gas 7 2 Existing utilities 1 0 Soft bottom in rock 2 2 Soft zones in rock 4 2 Hard, abrasive rock (TBMs) 5 2 Face instability, rock 5 1 Roof slabbing 4 1 Pressure binding (equipment) 4 4 Mucking 5 2 Surface subsidence 9 2 Face instability, soil 11 5 Obstructions (boulders, piles, 12 11 high rock in invert, cemented sand) Steering problems 4 0 Air slaking 1 0 *As noted earlier, in this report the word •claim• is a shorthand expression that encompasses all requests for extras as a result of an unexpected subsurface situation. The percentage of incidence as a problem indicated for groundwater does not account for its role as a contributor to the incidence or 96 

severity of other conditions (e.g., flowing ground, face instability), which would raise its rating significantly. As explained in Appendix c, there may have been instances where unclear original sources of informa- tion led the subcommittee to label occurrences of •flowing• ground (which is wet) as •running• ground (which is dry). This is one of the reasons that the problem tabulations do not give as much weight to groundwater as it deserves. The subcommittee believes that water plays a large and varied role in tunnel construction difficulties; yet it may not always appear in the simplified listing of problems encountered in a project because it is a secondary contributor to the primary problem. For example, face instability would in many cases not have been a prob- lem without the presence of water to reduce friction along joint surfaces or create seepage pressure, although the water might exist in quantities too small to deserve mention under groundwater inflow. In the same way, if only half of the recorded occurrences of running ground were really flowing ground, then that would raise by 12 the number of projects for which groundwater was a contributing cause of significant problems. Of all the problems, the highest incidence of claims (16 percent) was recorded for the grouping with the highest incidence of occurrence, i.e., blocky/slabby rock, overbreak, cave-ins. Of the other five con- ditions reported most frequently, three exhibit a similar relationship to claims. However, groundwater ranks several positions higher in problem incidence than in claim incidence; the ranking for obstructions is the reverse. Even so, the overall relationship is unchanged: the six conditions causing the most problems cause the most claims. Gener- ally thereafter, it is difficult to determine a pattern by ranking. However, the significance of a problem is related not only to fre- quency of occurrence but also to magnitude of impact. The tabulations in Table 7.3 can be translated to obtain a measure of impact by relating the incidence of claims to the occurence of problems. This method reveals that the relationship between occurrence and impact can be inversely proportional, as shown by the ratings presented in Table 7.4. For canparison purposes, Table 7.4 lists the problem conditions accord- ing to their frequency of occurrence, from highest to lowest. The impact rating is on an ascending scale, with a maximum value of 10. (For quick comparison between problem occurrence and significance, the impact rating can be multiplied by 10--i.e., a rating of 9.2 indicates a 92 percent incidence of claims per occurrence.) Table 7.4 reflects considerable impact--a rating higher than 6.5-- for six specific conditions, all of which occur infrequently: soft bottom in rock, pressure binding, obstructions, flowing ground, spalling and rock bursts, and noxious fluids. At this point the impact rating decreases suddenly, revealing six conditions closely grouped in the range from s.o to 4. 0. In this grouping, the impact rating is moder- ately high and the frequency is generally low, but a trend begins toward more direct proportion (i.e., for blocky/slabby rock and squeezing ground). For the remaining problems, the impact rating then drops to 3.3 (running ground), clusters again with five conditions (including groundwater inflow) between 2.8 and 1.8., and then falls to 0. 97 

TABLE 7.4 Impact Rating for Problem Conditions Conditions (t problems or occurrence) Impact Rating Blocky/slabby rock, overbreak, cave-ins (38) 4.2 Groundwater inflow (33) 1.8 Running ground (27) 3.3 Squeezing ground (19) 4.2 Obstructions (12) 9.2 (boulders, piles, high rock, cemented sand) Face instability, soil (11) 4.5 Surface subsidence (9) 2.2 Methane gas (7) 2.8 Noxious fluids (6) 6.6 Spalling, rock bursts (6) 6.6 Hard, abrasive rock, TBMs (5) 4.0 Face instability, rock (5) 2.0 Flowing ground (5) 8.o Mucking (5) 4.0 Pressure binding, equipment (4) 10.0 Roof slabbing (4) 2.5 Soft zones in rock (4) 5.0 Steering problems (4) 0 Soft bottom in rock (2) 10.0 Air slaking (1) 0 Existing utilities (1) 0 In certain instances (e.g., pressure binding), the severity of a problem can be linked with the sensitivity of the construction method to a particular condition. In others (e.g., overbreak, obstructions), it is less clear whether difficulty is more a function of the existing con- dition or the technique. It is likely that frequency of occurrence may sometimes be a moderating influence because it results in enhanced expe- rience and ability to cope that offsets the problem to some degree. In part, this may explain why the incidence of claims and/or impact rating for some problem conditions (e.g., runnning ground) is lower in relation to prevalence than might otherwise be expected. Unfortunately, several important aspects of the relationship between geotechnical conditions and construction problems cannot be readily dis- cerned or computed. They are the length of delays and degree of ineffi- ciencies that are introduced as a consequence, and their associated costs. The cost impact is not limited to construction dollars alone, but extends to project reliability and longevity. Although certain aspects remain ill defined, the plots and tabula- tions of numerical data yielded several interesting trends and quantita- tive values that help delineate the extent of the interaction of geology with construction and the effect of the geotechnical site investigation. The findings relate to the level of exploration, cost estimates, project costs, and claims. Before discussion of the findings, the manner of reporting the data merits a brief explanation. For the tabulated data, results are gen- erally presented as the arithmetic mean--referred to hereinafter as the • aver age, • in the conanonly understood sense of the word. For much of 98 

the plotted data, results are cited in terms of the •median.• Here, the value expressed by the median is considered more accurate than the every- day form of averaging, because the median accounts for the effects of significant skew in the samples. The techniques for deriving the values differ and, therefore, the terms •average• and •median• are not used in- terchangeably in this report. Level of Exploration In present practice, overall, the average number of boreholes drilled per 1,000 route ft of alignment is 2.4--i.e., a spacing that approaches 415 ft. It should be noted that these figures are based on 84 projects, of which 20 percent are in mountainous areas where tunnel depth can often exceed 1,000 ft and hole-to-hole spacing can reach thousands of feet. When data for these tunnels are excluded in order to reflect more common practice, then the average number of boreholes per 1,000 route ft nears 3.9, for a spacing that is about 260 ft. Although these tabulations provide a measure of exploration, they do not allow sufficiently for the effect of tunnel depth. Therefore, a more meaningful gauge can be obtained by determining the linear ft of borehole drilled per route ft of alignment. When the level of explora- tion is expressed in this manner, then the median lin ft of borehole per route ft is .34 in overall practice and .42 in common practice. (In this instance, the averages for lin ft per route ft are similar, .30 and .43, respectively.) Exploration costs were extremely difficult to compile because sep- arate records of the amounts spent were often not available or were incomplete. In addition, the task of apportioning costs for investiga- tion programs overlapping several projects was complex. As a result, the figures for exploration costs are considered less reliable than others reported herein. Of the 84 study projects, exploration costs for 36 were obtained. Information was sufficient for 30 projects (except as noted) to permit the extrapolations shown in Table 7. 5 Although some inconsistencies in matching samples were encountered, the preponderance of the data was ob- tained from projects for which figures were consistently available for each item tabulated. Therefore, the small variation in matching samples did not affect the results significantly. TABLE 7.5 Exploration Costs COmpared to Construction COsts Exploration Costs ($ millions) Construction Expressed as t Construction ($ millions) Total OVerall Range Median Engineer's Estimate 829.87* 9.80* 1.18 .01-24.4 .44 Basic COnstruction** 661.29 11.53 1.74 .02-17.5 .75 As COmpleted 694.00 11.53 1.66 .01-17.5 .70 *Figures are based on data for 28 projects. **Costs excluding claims awarded. 99 

Figure 7.1 illustrates more clearly the degree and nature of the scatter indicated by the range cited for exploration costs as a percent of construction costs. It is evident that funds expended for site inves- tigation programs do not rise with increasing project costs. Rather, a significant number of the more costly projects (i.e., those in the upper half of the scale) exhibit a decrease in exploration funds to a point well below the median. It is in the mid range that the number of proj- ects above the median generally equals the number below. However, only about 30 percent of these projects approach the median within a reason- ably small range of scatter. OVerall, these results indicate that present practice is to devote a relatively small portion of project costs to a site investigation pro- gram. In some instances, low expenditures may be warranted because a sufficient body of information may be available from explorations con- ducted for overlapping projects, or nearby projects, or from other sources such as aerial surveys and regional geologic reports. However, these circumstances cannot be assumed to explain entirely the general low level of expenditures or the scatter in the data. Even though the cost of construction is not always in direct proportion to the geotech- nical complexity or extent of a project, the relationship between these factors is obvious. On that basis, it is apparent that level of explo- ration costs does not correlate satisfactorily with construction costs. Estimates of Costa Engineer and Contractor The engineer •s estimate is a measurement of costs that is used by the owner for a variety of purposes throughout the conceptual to completion phases of a project. Essentially it serves as a benchmark for the devel- opment and evaluation of the components of the planning, design, bid- ding, and construction processes. As such, the engineer •s estimate is depended on to predict the actual project costs with reasonable accuracy. Figures 7.2, 7.3, and 7.4 compare the as-completed costs for mined tunnels with the engineer's estimate. The comparison examines the cost relationship in terms of several parameters representing the level of exploration. The individual results combine to form a more comprehen- sive basis for correlation. A review of Figures 7.2 through 7.4 reveals that as-completed costs differ significantly (±50 percent) from the engineer's estimate when the level of effort or funds devoted to geotechnical site investigations are low. However, this degree of variability is a reasonable occurrence only during the earlier stages of the initial conceptual work--i.e., when the exploration program is still in progress. This circumstance suggests that general exploration practice is providing inadequate in- formation for reliably estimating as-completed costs. The suitability of the site investigation also must be considered for its sensitivity to the effort level, an important concern because of its potential influ- ence on reliability. The deviation between as-completed and estimated costs decreases as exploration increases. Figure 7.2 indicates that the engineer's estimate becomes a more reliable tool for predicting actual costs when sufficient exploration has in fact been accomplished--i.e., boreholes at greater 100 

5.4 4B 9B 6.7 24.4 15.3 17 17.5 4.5 4.0 1- 3.5 Cl) 0 0 z 0 i= 3.0 0 :::> a:: t; z .... 0 0 u 2.5 .... u.. 0 ~ Cl) <( t-' 2.0 Cl) 0 0 z Q 1- <( a:: 0 ..J II. X w 1.0 - - - - - ENGINEER'S ESTIMATE ------~ AS COMPLETED - - - - - - - - - (NOT TO SCALE) INOT TO SCALE) FIGURE 7.1 • 

• 180 0 Selected Cue Studies 1, 2, 3, 4, 5, 6, 7, 8, 9 1601- • • Project Data Not Shown: X Axil 2.41 Y Axil 168.8 3.18 80.5 1401- • 3.61 4.55 66.0 208.2 w 1- < • 0 4.99 0 7.48 82.0 148.1 ~ 0 8.23 166.1 1- 0 9.75 324.9 Cl) w ~ a: w 0 • w z • 0 0 • • t.? z w 100 - - - - - · - . - - - - - - - - - - - ------ENGINEER'SESTIMATE-- ~------- .... 1.1.. 0 •• • • 0 N ?fl. Cl) 8 ••• • < ..... Cl) 0 0 0 w 80 • • , • 0 • • • • • • • 1- w ..J D. 60 • :::E 0 0 ~ 40 • 20 .34 Median, Overall .42 Median, Common 0~~--~--~~U-~---L--~--~~---L--~--~_..J~~---L--~--L-~---L--~ 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 BOREHOLES, IN LINEAR FT PER ROUTE FT OF TUNNEL ALIGNMENT FIGURE 7.2 

180 0 Selected c- Studies 1, 2, 3, 4, 7,8 Proj.:t Data Not Shown: XAxia Y Axia 017.0 148.1 0 w 1- 140 • 0 < ~ ~ w ~ a: w 0 w z • C) z w 100 t-- - - - - - . . 8 - - - - - - - - - - - - - - - - - - - - - - - - - - - ENGINEER'S ESTIMATE----- ~ c w u.. 0 ~ •• (/) < ~ •• • • 0 u 0 • 0 w 1- • • w ~ Q. • ~ 0 u ,;, < .75 Median, Overall 0~-L~~--~~--~--~~---L--~--~-L--~--~~--~--~~L--L--~~ 0 2 3 4 5 6 7 8 9 10 EXPLORATION COST, AS% OF BASIC CONSTRUCTION COST (EXCLUDES CHANGES AWARDED I FIGURE 7.3 

• 0 Selected Cae Studies 1, 2. 3, 4, 7. 8 Projct O.ta Not Shown: X Axis Y Axis 015.3 156.1 024.4 148.1 w .... • c( 0 :E ~ en w ~ a: w 0 w z Cl z • . w - - - - - - - - - - - - - · - - - - - - - - ---------ENGINEER'S E S T I M A T E - - - - - - - - - ..... c ~ 0 • ~ en • • c( ~ • • • 8 0 w 0 • • • .... _, w A. :E • 8 ~ 00 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 EXPLORATION COST, AS% OF ENGINEER'S ESTIMATE FIGURE 7.4 

than 0.6 linear ft per route ft. At this point, a substantial reduction is reflected in the frequency and degree of scatter above the estimate. A similar trend is exhibited in Figures 7.3 and 7.4, when funds expended exceed one percent. Decreasing scatter below the estimate is less marked in degree, but it is a tendency worth noting. The contractor's bid was examined in terms of two of the three param- eters used in evaluating the engineer's estimate. Appropriately, the major difference in approach was to review the relationship of the con- tractor's bid to both the engineer's estimate and as-completed costs. These results are presented in Figures 7.5, 7.6, and 7.7. In Figure 7.5 it is apparent that the discrepancy (.:!:sO percent) between the bid and engineer 1 s estimate decreases as the exploration level nears and continues beyond 0.6 linear ft of borehole per route ft of tunnel alignment. Then, the bid begins to approach, and is generally less than, the engineer's estimate. Interestingly, it can be observed that Figure 7. 5 reflects results distinctly similar to Figure 7. 2 with respect to incidence, degree, and pattern of scatter. At least a partial explanation for this similarity is provided by a review of Figures 7.6 and 7. 7, which compare the bid estimate and as- completed costs. Here, the incidence of scatter is consistent with that exhibited previously for low levels of exploration, but the degree of scatter is less pronounced (!30 percent rather than .:!:so percent). Moreover, at more suitable levels of exploration (greater than 0.6 for boreholes or one percent of the engineer's estimate), the convergence of the contractor's bid with as-completed costs is excellent, in terms both of degree and consistent pattern. The benefits resulting from the geo- technical site investigation are obvious. The difference in effects noted for the engineer's estimate and the contractor's bid merits attention to consider some of the possible causes. First, it might be expected that the contractor would be more experienced in evaluating requirements for tunnel construction suited to various purposes and ground conditions. Moreover, it is the contrac- tor's business to be accurate in determining the cost of individual elements so that an advantageous cash flow can be maintained. The margin between profit and loss is rarely sufficient to accommodate major inaccuracies without severe consequences. In comparison, the engineer's estimate is intended to predict total costs for the entire project (of which the mined tunnel is only a part) with a reasonable degree of accu- racy, which permits a more flexible approach. However, this built-in tolerance can be diminished or even eliminated if the estimating process is constrained. Among the elements that particularly influence the results are inflation before and during construction, constructibility, and detailed subsur- face information. I f policy, procedures, or cir- cumstances limit the determination or incorporation of any such basic components, then the accuracy of the engineer's estimate can be reduced accordingly and often to a profound degree. As inaccuracies escalate, it is increasingly difficult to avoid distortion of the estimate for the entire project. Certainly the bases for accuracy differ somewhat between contractor and owner, as well as the strictness of the criteria. In the final anal- ysis, only the owner can determine if the criteria have been satisfied when the engineer's estimate is simply higher than as-completed costs. 105 

180 • • 0 Selected a- Swdi• 1. 2. 3, 4, 5, 6. 7, a. 9 Project O.tll Not Shown: X Axis Y Axis 2.41 160.3 111) ,J 3.18 84.0 3.61 88.2 • 4 .56 4.99 206.1 81.7 w . 0 7.48 0 8.23 124.2 380.7 1- 9.75 83.5 c( ~ ~ ·- • en w en 120t-· ' • a: w w •• z 0 C) z 100 - - . - - - .....-- - - - - - - - - . . . - . - - - - - - - - - - - - ENGINEER'S ESTIMATE - ·---- .• ...... w ••• o • 0 u.. :· • • Cl'l 0 • •• • <It 0 ~ \ •• 801--. • 0 ci • • • • I (I) • • • • ~ a: ~ 0 c( a: 601-- ' 0 1- z 0 •• 0 .34 Medien, Owrell 0~~---L--~~~~~-L--~--~--L-~---L--~--L-~~-L--~--~~~~--~ 0 0.2 0.4 0.6 0.8 1.0 12 1.4 1.6 1.8 2.0 BOREHOLES, IN LINEAR FT PER ROUTE FT OF TUNNEL ALIGNMENT FIGURE 7.5 

180 0 Selected C.• Studies 1, 2, 3, 4, 5, 6, 7, 8, 9 HIO Project 011111 Not Shown: X Axis Y Axis 4.56 99.0 4.99 99.8 0 7.48 86.0 140 08.23 98.5 1- 09.75 25.7 Cl) 0 u 0 w • 1- w ••• •• _, • • ...J ~ ~ 0 u ~ \Ji • • • • lh 100 le, -..... , - - - . - . - - . - - - - - - - - - - - -AS.COMPLETEDS - - - - - - - - - - - - 4( ~ 0 ...a u. 0 ~ •a• • e • • • • •• • Cl) 4( ci • 0 iii Cl) • 0 iE: 0 0 • 1- u 4( a: 1- z 8 .34 Medilln, Ownlll .42 Median, Common 0~--~~-'---~---L---L--~--~~--L---~---'----L---L--~--~----~--' 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 BOREHOLES, IN LINEAR FT PER ROUTE FT OF TUNNEL ALIGNMENT FIGURE 7.6 

180 0 Selected C.. Studies 1, 2, 3, 4, 7, 8 160 Project Oete Not Shown: X Axis Y Axis 024.4 86.0 0 16.3 98.6 140 ~ ~ c.> 0 w 120 ~ w ...I CL. ~ 0 c.> • • ~-#-.-----.-------------.----------------AS-COMPLETED$ • ~ 100 .... c Q) u.. 0 ~ • a • (I) < 80 0 0 iD (I) 0 a: 0 ~ 0 c.> eo < a: ~ z 8 40 .44 Medien, Owrell 20 o~~--~--L-~---L--~-....1~-L--~--~~--~--.__. ___ L_ _~__,~~--~--~_.--~ 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 EXPLORATION COST, AS 'Ko OF ENGINEER'S ESTIMATE FIGURE 7.7 

Claias Requests for extra payments for unexpected subsurface situations--i.e., claims--appear to be a significant part of tunnel cost. Of the 84 mined tunnels studied, 49 reported claims related to geologic and/or subsur- face conditions. (Three of these claims were for comparatively small amounts, i.e., less than $25,000.) Projects without claims totaled 29, and there were 6 for which no information was available on whether claims had occurred or not. overall, then, it appears that about 60 percent* of tunnel projects entail claims, and that claims are substantial for about 55 percent of tunnel projects. Stated differently, of the proj- ects experiencing claims, nearly 95 percent of the claims are for large amounts. Of the 49 tunnels with claims, there were 32 for which sufficient data were reported (Table 7.6) to permit evaluation. Combining the fig- ures in Table 7.6 yields the following sums (in constant 1982 dollars): total claims of $253.7 million, claims paid of $161.8 million, basic con- struction costs of $1,364.8 million, and as-completed costs of $1,526.6 million. Examination of the data for the 32 tunnels reveals that 11 of the claims (8 major and 3 minor) were settled for essentially 100 percent of the amounts requested, one major claim was settled for 115 percent, and 3 major claims were settled for zero payment. The remaining 17 claims were settled for sums varying from 10 to 70 percent of the amount re- quested, for an average of 39 percent. There is no apparent relationship between the amount of the claim--made or paid--and the original size of the project. overall, the indication is that payments were settled at about 64 percent of the original total claimed. These payments amounted to nearly 12 percent of the basic construction costs. (If this average were com- puted from the claimant's viewpoint--i.e., increase the as-completed total by $91.9 million [the difference between claims made and paid) to reflect construction costs considered justifiable--the settled payment would approach 10 percent.) In view of these results, the possible influence of the exploration program is an unquestionably relevant concern. Therefore, claims were reviewed in terms of several parameters that served for correlation in the preceding sections. To obtain an appropriate comparison, the data base was expanded by removing the limitations imposed for Table 7 .6. Claims made were examined in terms of both the engineer's estimate and contractor's bid, and then compared with the exploration level for bore- holes. *It is possible that this figure may be low as an extrapolation to industry-wide occurrence. At least two owners who volunteered completed projects are known to have each withheld one newly completed project with claims currently in litigation. If very many of the participating owners faced similar dilemmas, then the sample may be biased toward the •no claims• end of the study spectrum. Also, some projects with litigation completed may have been withheld to avoid possible embarrassment to any interested parties. The subcommittee hopes that the potential for dis- tortion is minimal and limited to unresolved claims. 109 

TABLE 7.6 Comparison of Claims with Construction Costs Totals* ae22rted for Claims Totals* for Construction Claims Made Claims· Paid Basic** AB-com21eted 0.008 0.008 4.69 4.7 0.008 o.ooa 12.59 12.6 0.010 0.010 4.79 4.8 0.060 0.020 10.18 10.2 0.250 0.150 11.65 11.8 0.350 0.350 16.15 16.5 0.600 0.600 14.4 15.0 0.680 0.170 16.43 16.6 0.700 0.700 21.4 22.1 0.747+ 0.747+ 32.3 33.05 0.787 0.387 25.013 25.4 o.8oo 0 5.9 5.9 1.000 0.200 15.2 15.4 1.290 0.370 28.83 29.2 1.400 0 19.4 19.4 1.600+ 1.600 29.0 30.6 1.800+ 1.800 99.9 101.7 2.000 0 275.5 275.5 2.100 2.100 51.2 53.3 2.100 1.200 71.5 72.7 2.600 1.500 32.3 33.8 5.400 2.000 18.5 20.5 6.900 1.300 29.7 31.0 7.500 0.750 19.55 20.3 8.100 2.980 73.22 76.2 11.800 3.400 42.6 46.0 19.200 11.100 104.5 115.6 25.100 7.950 140.55 148.5 25.100+ 25.100 44.2 69.3 28.200 20.200 34.8 55.0 37.000 7.500 29.8 37.3 58.500 67.600 29.1 96.7 *Constant 1982 dollars, in millions. **Costs excluding claims paid. Results of the comparison, presented in Figure 7.8, indicate a well defined relationship between claims and the exploration effort. At low levels of exploration, approximately 50 percent of the requests were for amounts greater than 10 percent of the engineer's and contractor's esti- mates. OVerall, claims averaged 29 percent of the engineer •s estimate and close to 28 percent of the contractor's bid. As soon as exploration exceeds 0.6 linear ft of borehole per route ft of alignment, a marked decrease occurs in the number and size of the claims. This trend down- ward continues sharply as the borehole level exploration increases. A similar comparison of claims and funds expended for exploration produced matching results. Thus, it is clear that the site investigation program can moderate the occurrence of claims and their severity. 110 

80 1 • Bid O.te Only Aveileble • - Selected C.. Studiel 2, 3, 6, 7, 8, 9 70 l Project O.• Nol Shown: X Axil OY Axil e Y Axil • 0.01 - 9.75 188.1 197.8 139.7 235.0 w - 0.41 72.6 126.9 1- <to :1- 60 -CD lntn wcr 0 cnO • 1- cru We( wcr ~~-- • 0 ~z 50 0 zo wo ILLL 00 #lit ~~ oe 40 • 0 w 1- Cl) w :::> 0 30 w cr in w • 8 ~ z t 4( X p 20 10 • & 8 8 •• 0 0 0 ... •• •pp a~ • 8 • • 0 0 3 8 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 BOREHOLES, IN LINEAR FT PER ROUTE FT OF TUNNEL ALIGNMENT FIGURE 7.8 111 

Project Costa The ultimate cost of projects can be estimated accurately and controlled or moderated without sacrificing fair compensation. These goals are at- tainable when the site investigation program is conducted at a sufficient level to permit thorough evaluation of the subsurface by all parties to the construction process and according to their specific needs. However, the exploration program can only contribute successfully to these goals if the level of effort is increased as a matter of general practice. This belief is based in part on the following findings, which relate to the relationship between exploration and project problems and costs. • Of all the projects studied, only 11 reported no significant prob- lems and none reported minor problems alone. For more than 85 percent of construction projects, the typical level of site investigation is too low to characterize subsurface conditions adequately in order to plan for or avoid impact on constructibility. This circumstance leads to inaccurate budgets and schedules, inappropriate construction procedures, unnecessarily (often) high increases in as-completed costs, claiJDS and litigation, and difficulties with operations and maintenance. • The engineer's estimate varies ±so percent from both bid costs and as-completed costs at the levels of exploration commonly practiced. These deviations are a source of uncertainty for the owner, contractor, and the public, and promote costly adversary relationships. When typical exploration practice is increased, bids and as-completed costs tend to equal or even fall below the engineer's estimate. • Claims related to unanticipated subsurface conditions occur for about 60 percent of construction projects. In some instances, a claim may be part of the fair cost of a project. Overall, however, claims and disputes result in inefficiencies that are expensive for the owner and contractor. Improving site characterization by increasing exploration reduces the incidence of claims and attendant effects; when unnecessary claims are avoided, construction is more economical. • A significant portion of project costs stems from claims settle- ments rather than from investigation of the site for design and con- struction purposes. The typical one percent (about) of project costs expended for exploration is obviously too low when compared with the average 12 percent of project costs devoted to settled payments for claims. If this figure were extrapolated from 65 to 100 percent of the projects reporting claims, then the average minimum settlement is slightly greater than 7 percent of project costs. However, these aver- ages represent only the amounts publicly paid in settlement of a claim. For the adversaries, there are additional costs for staff and legal ser- vices that usually are not disclosed. These •hidden• costs can be tre- mendously high, and in some instances may reach a total that represents a significant portion of the project cost The geotechnical site investigation cannot predict every problem that may be encountered, and attempts to do so generally result in pro- grams that are disproportionately expensive for the value received. For every underground project, cost-benefit is a key element. Increasing the level of effort and funds for exploration is demonstrably beneficial and cost-effective. 112