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.

6. Selected Case Studies The ten projects (nine mined tunnels and one shaft) presented in this section were selected because they represent problems or situations which the subcommittee feels it will be instructive to explore in de- tail. The case studies of the mined tunnel projects were chosen to match as many of the following criteria as possible: • Taken together, the projects should represent the widest possible range of basic problems encountered, as reflected under that subheading in the project abstract. (In fact, most of the selected cases will illustrate two or three major problem groups and twice as many subgroups.) • The nine cases should represent at least several different tunnel purposes, such as water conveyance, power generation, rapid transit, etc. • Each case must be based on a thoroughly researched study proj- ect. This eliminated, for example, those projects for which a follow-up interview with the owner was not carried out. • Each case must be based on a study project for which all con- struction has been completed. • Each case must be based on a study project for which all lit- igation (if any) has been resolved. • At least one of the cases should be based on a project with no significant construction problems and no subsurface-related cost overruns. Although all of the 84 mined tunnel projects illustrated some prob- lem or feature that might deserve discussion, the 9 cases selected best met the widest range of stated criteria. The projects represent only 6 of the 28 owners or agencies who provided information for the study. Thus, it might appear that those six were singled out for particular criticism, but that would be a misconception. It is purely coincidence, and not perceived flaws in philosophy of design or site investigation, that caused the selected case studies to represent so few of the owners or agencies. In any case, limiting the number of projects selected for special examination necessarily restricted the set of owners and agen- cies. For the shaft case study, the choice was much more severely limited because only three deep shaft projects were studied. The subcommittee 46 

decided to use the Waste Isolation Pilot Plant (WIPP) project because it is comparable to the type of undertaking contemplated in the construc- tion of waste repositories requiring a number of deep shafts for access to chambers designed for storage of radioactive and other hazardous sub- stances. Neither the Loon Lake penstock shaft nor the Brunswick No. 3 mine shaft could yield the maximum amount of information to the parties involved in the planning for deep underground storage, because of their different needs and opportunities for subsurface investigations. There- fore, the WIPP project was selected as best meeting the criterion of ap- plicability to user needs. It should be noted that the costs presented in the case studies are as taken from bid tabulations and pay vouchers. The dollars represent values for the years in which they were obligated or paid, with no esca- lation factors applied. 47 

CASE STUDY NO. 1 Name of Project& MBTA Red Line Extension, Porter Square Station Purpose& Passenger station for subway system Locations Massachusetts (Cambridge) Construction Period& March 1980--June 1981 Site Investigation Period& 1976--1978 Sizes Trainroom 490 ft longJ 45 ft 7 in. high by 70 ft 6 in. wide. Crossover 68 ft longJ 37 ft 1 in. high by 44 ft 2 in. wide. Project Cost& Estimated $36,969,138 Bid $43,887,900 As Completed $44,877,854 (includes all extra payments) Mined Tunnel Construction Costs Estimated $13,035,444 Bid $21,045,650 General Contract Mods -$701,598 Subsurface Related Overruns $0 As Completed $20,344,052 Subsurface Investigation Costs $2,000,000 (plus or minus) Summary of Site Geologys Predominantly fresh to slightly weathered, bedded argillite with a slight dip, and overlain by thin glacial till, marine clay, outwash sands, and miscellaneous surficial till. Minor in- truded dikes of basalt and andesite. RQDs indicate generally fair to excellent quality, but two faults were identified in addition to fre- quent shears perpendicular to the station axis. Unconfined compressive strengths were 9, 740 to 45,500 psi for argillite and 15,900 to 24,800 psi for igneous rocks. Joints and fractures, a source of stored water, were mostly tight but areas adjacent to intrusive dikes likely to be more pervious. Depth of overburden ranged from 64 to 82 ft above tunnel crown (30 to 47 ft of rock cover above crown). Static water table at 15 to 20 ft below surfaceJ no water inflow predicted. Design Criterias Maximum total load of 8,800 psf for final liningJ concentrated rock loads of 1,200 to 3,500 psf also used for other geome- tries. Water level at El. 123 (60 ft above crown in trainroom and 48 ft above crown in crossover chamber). Contract Provisionas Types Unit price per cubic yard of excavation and per unit of lining components (support steel and shotcrete). 48 

Stipulations & Schedule and/or time of completion& Total contract to be completed by 9/11/82. Definition of delay and suspension of work. Liquidated damages: $2,500 per calendar day of delay. Payment: Monthly; 5% retainage (may be eliminated after SO\ comple- tion). Construction method: Drill-and-blast (4-stage scheme, modifiable by contractor). Restrictions: Work not permitted on weekends or holidays without approval. Surface hauling not allowed between 11&00 p.m. and 7:00 a.m.; route and disposal site specified. Monitoring required for blasting. Strict noise level control. Disputes resolution: Decision by owner's engineer. If agreement is reached, contractor reimbursed at cost plus U, 6%, or 10%, as determined by engineer. Recourse is appeal to director of con- struction, then a review board, then litigation. Geotechnical data made part of contract documents& •Geotechnical Interpretive Report• (available for purchase), which included cross sections and test data. Boring logs and pilot tunnel maps included in contract drawings. Core samples available for inspection by ap- pointment. Disclaiaera& None with respect to owner-furnished information on subsurface conditions. Changed-conditions clause& Yes COnstruction Method& Drill-and-blast, 3-stage excavation (top head- ing, intermediate heading, and lower bench). Primary support of steel ribs, rock bolts, and 3 stages of shotcrete. Permanent support the same as primary support plus 4th stage of shotcrete (minimum total thickness of 15 in.). COnditions Encountered& Relatively good conditions, essentially as predicted and perhaps slightly better. The contract was modified to permit the contractor to change from a 4-stage to a 3-stage excavation scheme. Groundwater inflow of 42 gpm for two months, until underground reservoir drained. Problems Encountered& Construction& None of major consequence. There was a delay of perhaps three weeks when a fault was encountered in the portal. (This was early in the learning curve of perfecting the support system.) Operations and Maintenance& Groundwater flowing through the bed rock has sufficient concentration of caco 3 to be considered cal- careous. There is evidence that the carbonates may be precipitating in open air, enough to begin clogging drainage systems over a period 49 

riod of time. At this writing, it is not known how serious the problem may become. (The problem is well documented on several sections of the washington Metro system.) Resolution of Assertions Re Subsurface Changes a NO assertions were made with respect to subsurface changes. Analysis/Opinions When planning for the Porter Square site investiga- tion was begun, just after the mid 1970s, there had been no previous ex- perience with design and construction of a large shallow chamber in the argillites around Boston. The WMATA system in Washington, D.C., had been providing a record of experience since the late 1960s, but those shallow chambers were constructed in schists and gneisses of a complete- ly different geological regime. The MBTA geotechnical engineer appar- ently decided that a very great deal of information about the rock in his local area would have to be developed before attempting such excava- tions and therefore took a very conservative approach to the site inves- tigation. The resulting body of knowledge was quite impressive and was undoubtedly a major factor in the absence of cost overruns in the mined opening. Because the investigation seems to have been extremely suc- cessful in achieving its primary purpose, cost effectiveness is the only aspect of the program that is legitimately open to debate. Shallow rock chambers are generally regarded as some of the more critical of the civil engineering projects because of the excavation spans involved, the probability of closely spaced discontinuities (and perhaps intense weathering) so near the bed rock surface, and the general looseness of rock blocks because gravity induced stresses are too low to keep them pressed firmly together. An absolute minimum site investigation for such construction would certainly include a generous number of boreholes with rock coring, lab testing to determine strength, hardness, etc., and borehole water level and permeability tests. Prudence would dictate the use of a few oriented core holes to determine rock structure attitude and maybe some overcoring tests for quantifying and orienting locked-in stresses. It would not be unreasonable to consider a small pilot tunnel for detailed mapping and later access by bidders. Perhaps in addition to or as a substitute for some of the above, one might consider pumping tests, blast vibration tests, or the construction of a test shaft. The interesting thing about the Porter Square investigation is that it encompassed all of the above techniques of rock and soil exploration. Although some of the tabulated costs are estimates or bid prices rather than final recorded figures, it appears that the total amount spent for the complete program was in the neighborhood of $2 million. With the final cost of the mined station chamber being about $20.3 million, a best guess is that the owner's exploration costs were about 9.8 percent of the construction costs (ignoring the fact that exploration dollars had a mid-to-late 1970s value while construction dollars had an early 1980s value) • Still another way of looking at the matter is to note that the owner originally estimated the cost of the mined opening at slightly more than $13 million. Hence, exploration costs were perhaps 15.3 per~ent of the presumed construction costs (again ignoring the ef- fects of inflation). so 

It is logical to ask if the scope of the site investigation could not have been;reduced without detracting too much from the data base de- veloped for designers and bidders. Because the pilot tunnel (excavation bid price of $1,683,800) took the lion's share of the exploration bud- get, a closer look at its cost effectiveness seems warranted. Sized at 12 ft by 12 ft, this opening certainly did its job of providing an op- portunity for measuring water infiltration, confirming rock joint pat- terns and conditions, and demonstrating how certain joint sets would control overbreak. It also was instrumental in locating two small faults and two minor igneous dikes that had been missed by core bor- ings. However, water infiltration had already been measured with accu- racy in an inspection shaft (excavation bid price of $69,070) that was 36 in. in diameter and lll. 5 ft deep. Assuming the pilot tunnel was truly needed to confirm the other geologic features, it could have been done just as easily in a smaller tunnel, perhaps 6 ft wide by 8 ft high. The smaller size surely would have cut the cost of the opening and would not have provided so much opportunity for the rock in the crown to loosen prior to opening up the full station chamber. The argument that only a large pilot tunnel easily permits the early installation of rock dowels for station excavation support may be a case of circular reason- ing because too large an opening can be the very cause for needing such dowels in the first place. Indeed, a small construction problem did de- velop at Porter Square because blasting for the pilot tunnel damaged the integrity of the rock enough to require shotcreting of the pilot tunnel roof ahead of the advancing station chamber top heading in order to keep down overbreak. One may say that this is all quibbling and the only important fact is that the pilot tunnel (in conjunction with the other elements of the site investigation) obviously reassured bidders about conditions, mini- mized construction problems, and eliminated cost overruns, thereby pay- ing for itself in the long run. The only easy way to make a tentative judgment on this is to look at the construction costs, which break down as follows: Mined Tunnel Total Contract Engineer's Estimate $13,035,444 $36,969,138 Low Bid 21,045,650 43,887,900 Contract Modifications -701,598 +989,954 Geology Related Claims 0 0 $20,344,052 $44,877,854 It is true that i f one compares the low bid amount with the final cost figures, there were no geology related overruns in the station chamber. There was even an apparent savings, the exact reason for which was never made clear to the subcommittee interviewer. However, the bot- tom line is that the low bid and the final costs came in at approxi- mately $7 million more than the owner had expected to pay. In a compet- itive situation, the question to be asked is whether less subsurface information from a less expensive exploration program would have raised the bid price by any substantial percentage of the $2 million (plus or minus) that was spent. This leads to the question of whether a less in- formed contractor might have encountered enough construction surprises to raise the ultimate cost to any great degree. There is no way to pro- 51 

duce any adequate proof when speculating on •what might have been,• but cutting the cost of the site investigation in half would have netted the owner approximately $1 million in early money savings to balance against bidding contingencies and potential construction cost overruns. In developing subsurface information, one must always aska •At what expenditure level do exploration costs begin to exceed potential construction savings?• No amount of money spent on exploration can re- move all construction uncertainty, so the owner and the geotechnical engineer must draw the line at some point. This project may be an exaa- ple of one where a line was drawn slightly beyond the bounds of cost ef- fectiveness. 52 

CASE STUDY HO. 2 Name of Projecta WMATA Section C-4, Huntington Route (Contract 1C0041) Purpoaea Running tunnels for subway system Locationa Washington, D.C. (northwest quadrant and under the Potomac River) Construction Perioda November 1972--August 1973 (shield tunnels) Site Investigation Perioda September 1966--August 1969 Sizea Soft Ground 2,740 ft longJ 20 ft 6 in. diameter. Mixed Face 1,069 ft long; 19 ft 8 in. diameter. Rock 8,303 ft long, 19 ft 8 in. high by 19 ft 8 in. wide Project Costa Estimated $26,930,647 Bid $23,397,053 As Completed $32,009,752 (includes all extra payments) Mined Tunnel Construction Costa Estimated $18,230,267 Bid $15,649,372 General Contract Mods $99,788 Subsurface Related overruns $9,217,999 As Completed $24,967,159 Subsurface Investigation Costa $98,150 pre-bid Summary of Site Geologya Recent alluvium and man-made fill overlying Pleistocene terrace deposits (fine and coarse grained sediments) over- lying decomposed rock and schistose gneiss bed rock. Eastern portion of alignment in terrace sands and gravels with boulders near base of depos- it and layers of clayey silt and silty clay throughout the upper reaches. A relatively thin layer of saprolitic decomposed rock separates the ter- race deposits from the underlying bed rock. Most of tunnel beneath the Potomac River in quartz-mica schist-to-gneiss of the Wissahickon and Sykesville formations. Foliation not particularly pronounced but shear zones common. Rock quality highly variable, ranging from slightly to highly jointed, with talc coating on some joint surfaces. Slightly to highly weathered, with some weathering zones at depth beneath sound rock. Unconfined compressive strength varying between 560 psi (in weathered zones) and 15,860 psi. Overburden ranges from 12 to 80 ft above the crown; soil thickness ranges from 0 to 50 ft except much thicker (120 ft~ in gorge on east side of the river. Median permeability was 4 x 10 in rock. Predicted water inflow of 7 gpm in rock. Design Criteriaa Water pressure (range) from 8 ft below the tunnel crown to 65 ft above the crown. 53 

Contract Provisions& Typea Unit price per linear ft of tunnel excavation as follows: 2,740 ft earth tunnel; 1,069 ft mixed face; 8,303 ft rock. (Total tunnel excavation length • 12,ll2 ft along 6,056 ft of alignment.) Unit prices for support items: shotcrete (cubic yard), ribs (each), steel (round). Estimated quantity variation limits set at 1St, without contract price adjustment. Stipulations a Schedule and/or time of completion: 730 days to complete tunnels. (Contractor to submit schedule, which then became the contract time.) Definition of delay and suspension of work. Liquidated damages: $1,500 per day of delay. Payment: monthly; lOt retainage (after SOt completion, may be re- duced at contracting officer's option). Construction method: TBM or drill-and-blast for rock tunnel. Vari- ous liner options, including shotcrete, cast-in-place concrete, and liner plates. Also option for either steel ribs or shotcrete and ribs in rock tunnels. Restrictions: Three shifts to be maintained when using a shield. Blasting not allowed from 10:30 p.m. to 7:00 a.m. in washington, D.c., or Virginia, but no restrictions in tunnels under the Po- tomac River. Hauling subject to local jurisdictions. Disputes resolution: Decision by owner's contracting officer. De- cision can be appealed within 30 days to owner's board of direc- tors; board decision final unless question is one of law that results in litigation. Geotechnical data made part of contract documents& Boring logs (bound directly into the contract documents). Core samples avail- able for inspection. (Subsurface investigation reports, including profiles and laboratory test data, available for inspection and copies could be obtained from the National Technical Information Service.) Disclaiaersa Yes; data presented for information only with dis- claimer on accuracy, interpretations, and conclusions in reports. Changed-conditions clausea Yes Construction Method a Dr ill-and-blast (boom mounted 4-dr ill jumbo) for rock tunnels and some mixed face. Shield in earth tunnels and some mixed face where rock was below springline. Primary support of steel sets, some shotcrete (initial portion of rock tunnels), and some spiling (soil roof of mixed-face tunnel) • Final lining of reinforced concrete (12 in.). Conditions Encountered& In soft ground and mixed face, essentially as predicted, except elevation of rock line higher than expected. Blocky conditions and excessive overbreak in rock, but this is a controversial 54 

matter (conditions varying from poor to fairly good, and probably no worse than predicted by owner). Rock between tunnel crown and Potomac River bottom possibly sounder than expectedJ water pumped for the dura- tion of the project was on the order of 50 million gallons, only 10 per- cent of the specified allowance. Problema Encountered& Construction& In soft ground, runs into the heading caused ground settlements, including two surface slumps. In mixed face, a higher than anticipated rock line for part of the extent resulted in a change from shield excavation to heading and bench. In rock, blockiness and overbreak resulted in the use of steel ribs rather than the design support system of rock bolts and shotcrete. Operations and Maintenance& At present, problems caused by ground conditions are minimal. Groundwater leakage is minor and there is hardly any buildup of the calcium carbonate precipitates that have plagued many other Metro rock tunnels. Drains were flushed perhaps 8 months ago (counting from January 1983) and still appear to be in decent condition. There is a bit of silt buildup in the drains at the low point of the tunnelsJ i t is not known whether the silt originates in construction debris or in joint fillings in the surrounding rock. There was a short-term maintenance problem that stemmed from a man-made condition. During construction, the tunnels penetrated soft ground saturated with a heavy, tar-like substance left from an old factory site. After tunnel completion the material continued to seep through the permanent concrete linings. Although not a fire hazard, it was messy and was carried by the drainage system to the pumping station beneath the Potomac River. When released into the river, the petrochemical was considered a minor environmental problem. The substance disappeared after a few years, possibly because the pocket was effectively drained. Resolution of Assertions Re Subeurface Changes& The contractor as- serted that he encountered higher rock than could be anticipated from the pre-bid data, primarily because the geologic profile contained a plotting error indicating a 1.5-ft higher top by scaling than by written dimension. The contractor had scaled dimensions from the profile to prepare the excavation bid estimate and maintained that the error had increased his excavation coats by a factor of four. The owner's consul- tants contended that the plotting error was minimal, that all other drawings were accurate, and the written dimensions should have taken precedence. In addition, the geotechnical report indicated that varia- tions of 2 to 5 ft in rock elevation could normally be expected. A claim was filed but settled prior to hearing at a cost of $162,788 (part of a blanket settlement). The contractor asserted that steel ribs on 2-ft centers had been required due to blocky ground and safety of excavation and personnel, maintaining essentially that the design support system of rock bolts and shotcrete was faulty and not sufficiently conservative. The owner 55 

disagreed, indicating (1) that the contractor had never attempted to construct the tunnel as designed or as bid, (2) that numerous ribs in- stalled evidenced no blocking, no loading, and no deformation, and (3) that the design support system could have been used effectively. Claims pertaining to overrun in ribs were settled during performance for $2,503,815 by owner's final decision. Claims made by the contractor totaled $12,768,374. Some were set- tled by owner decision without litigation, others were filed before the Corps of Engineers Board of Contract Appeals but settled before an actu- al hearing. The final amount awarded to the contractor was $9,217,999. Analysis/Opinion& WMATA's C-4 contract provides examples of two caa- pletely different kinds of subsurface problema that led to complications during construction. The first, caused by a higher rock line than the contractor apparently had a right to expect, is extremely common wherever a mined tunnel impinges on top of bed rock. It was recognized that the tunnels would transition from soft ground to mixed face to rock, and the contractor laid plans to push with his shield to the point where the rising rock would force him to abandon this method. However, due to an owner plotting error on one contract drawing and some rather simplistic borehole-to-borehole rock line projections by the contractor, the top of rock rose to a higher than expected elevation in the soft ground tunnels and slowed progress considerably. Probably contributing to the problem was the somewhat less than desirable borehole coverage, with spacings on an average of perhaps 200 ft apart and staggered from one side of the alignment to another. WMATA's present practice in similar circuastances is to make borings or pairs of borings (one on each side of the align- ment) on so- to 150-ft centers, coverage that is three to four times as tight as that provided on Section C-4. This constitutes acknowledgment of the fact that an owner can seldom go too far in determining the rock line when its presence is likely to affect a mined tunnel. In all fairness, however, it is difficult to say whether knowing the location of the top of rock with great precision would have made much difference in the ultimate cost of these particular tunnels . sec- tion C-4 was not designed to skim the top of rock in order to avoid mixed face conditions, it had to traverse those conditions in order to dive into rock, and a knowledge of their limits would not have lessened their extent or severity. The contractor made a high rock claim of $1,187,200 and ended up collecting $162,788 for it. A very precise knowledge of rock elevations would presumably have driven his bid up by a similar amount, and therefore it may be that the only money really •lost• was some relatively minor amount caused by the surprise factor and whatever the situation may have contributed to litigation expenses. By far the more serious of the C-4 problems was the one relating to rock conditions and how they affected tunnel support. The situation was quite complex, with many overlapping claims and counter claims which, had they been paid in full, would have netted the contractor extras worth +$12 million, however, they were finally settled for +$9 million. Though difficult to summarize without sacrificing accuracy, the basic problem appears to be that the contractor bid a construction option which he later decided was impossible to pursue. Passing up the chance to use a · TBM, he chose drill-and-blast tunnels with a mostly rock bolt 56 

and shotcrete lining, but with a designed support system of steel ribs on 4-ft centers in known weathered areas and shear zones. He then pro- ceeded to line the tunnels with steel ribs on 2-ft centers and supple- mented them with large amounts of miscellaneous steel, saying that the rock was obviously too poor throughout to be supported by the owner's design lining. The installation of so much tunnel steel interfered with shotcrete placement and mandated a reversion to cast-in-place concrete, a change which contributed heavily to the overruns. To an outsider, all indications are that the rock probably was no worse than envisioned by the owner all along, his site investigation had predicted conditions with relatively fair accuracy. We believe the ba- sic problem lay with contracting procedures and the way design and geo- logical information was passed along to bidders. The contract placed a great deal of responsibility for tunnel safety on the contractor and then presented him with a shotcrete and rock bolt support system with which few Americans had much experience in 1972 and which may have appeared to be on the outer limits of feasible technology. The fact is that, even then, tunnels were being supported by shotcrete in ground that was certainly no better than Section c-4. However, a mere design drawing probably constitutes little reassurance for a builder contem- plating relatively new support techniques. The C-4 contractor never tried the owner's design lining (which made it difficult to finally de- termine whether the ground was as predicted or not), but he might have been more willing if there had been a mechanism for explaining it to him. In the early 1970s, WHATA made its interpretive •subsurface inves- tigation reports• available for reading by bidders, but disclaimed any responsibility for conclusions drawn therefrom. In spite of this dis- claimer, the C-4 contractor did depend on the reports in putting to- gether a bid and at least had access to an accurate assessment of actual ground conditions. Unfortunately, but of necessity, a subsurface inves- tigation report is compiled during the early to mid stages of design. Therefore, it is often impossible for such a document to treat or com- ment on many important design and construction matters because they are not worked out until a later stage of development. This is especially true in the case of innovative ideas for design or construction. Hence, the C-4 contractor had no easy way of comprehending the rationale for a shotcrete and rock bolt support system in these particular tunnels, and this may have contributed to his unwillingness to give it a fair trial. Since 1975, WMATA has had a mechanism for passing along such im- portant information to bidders and construction managers for all mined tunnel projects. That mechanism is a report entitled •Geotechnical Ba- sis of Design and Construction Specifications,• or •Geotechnical Design Report• for short. This report is compiled by the tunnel designer and bound as an appendix into the construction specifications so there can be no doubt about its status as a contract document. The geotechnical design report sums up the important geologic information from the sub- surface investigation reports and then explains how the geology affected design and how it is likely to affect construction operations. Thus, the contractor and the resident engineer are fuily apprised of the de- signer's and the geotechnical engineer's intentions and advice and, as a result, the field work proceeds more smoothly than it otherwise would. 57 

Had there been such a report in the C-4 contract documents, it seeiiB likely that much misunderstanding and litigation would have been avoided. Aside from high rock and general rock conditions, a third and very minor C-4 problem is worth mentioning because it is symptomatic of the kind of occurrence that has proven more significant on other projects. The tar-like substance encountered beneath an old factory site was not volatile enough to be a true construction hazard and its general messi- ness is no longer very troublesome now that the pool has apparently drained. Nevertheless, the presence of this substance should not have been overlooked in the site investigation. 'l'o miss such a relatively innocuous substance means that one with greater potential for harm could have been missed just as easily. Many urban areas are dotted with spills from gasoline stations, factories, and the like, and it is in- cumbent upon investigators to identify such areas before the tunneler arrives to discover them for himself. sa 

CASB STUDY NO. 3 Halle of Projecta WMATA Section G-2 (Contract lG0021) Purpoaea Running tunnels and station for subway system Location& Washington, D.C. (northeast quadrant) Construction Period& October 1975--June 1978 Site Investigation Perioda April 1972--April 1975 Sizea 13,700 ft long, 20 ft 11 in. diameter. Project Costa Estimated $49,587,227 Bid $42,266,620 As Completed $48,555,357 (includes all extra payments) Mined Tunnel Construction Costa Estimated $31,831,000 Bid $18,226,940 General Contract Mods $86,204 Subsurface Related overruns $4,718,311 As COmpleted $23,031,455 Subsurface Investigation Costa $49,775 pre-bid S...ary of Site Geology& Stiff to hard Cretaceous plastic clays and sandy clays and compact to very compact silty sands, with many intermix- ing& and interlayering& of the three basic strata. OVerlain by Pleisto- cene terrace deposits and man-made fill. Depth of overburden ranged from 27 ft to 96 ft. Water table at 15 to 45 ft above tunnel crown. Median permeability was 4 x lo-6 fpmJ the highest measured 6 x lo-4 fpm. Tunneled soils stiff/compact due to preconsolidation. Silty sand often with less than 10 percent fines, making it unstable, especially where water difficult to draw down due to interfering clay lenses. Wettest material was the the clean sand lenses occurring in otherwise silty and clayey strata. Evidence of perched water due to pumping from household wells at depth while upper strata recharged by infiltration from the surface. Predominance of clayey materials hinders vertical movement of water. Design Criteria& Between 6 and 13 kips overburden load at tunnel springline: 15 to 45 ft head of water above crown. Contract Provisions& Typea Unit price per linear ft of earth tunnel and lining. Es- timated quantity variation limit set at 1St, without price adjust- ment. 59 

Stipulations& Schedule and/or time of completion: 910 calendar days for the total contract. (The contractor was required to submit for approval a detailed Logic Network Analysis with estimated activity durations and milestones for various major features, including running tun- nels estimated at 203 calendar days for mining.) Definition of delay and suspension of work. Liquidated damages: $2,500/day for certain specific features~ $1,500/day for the total contract. Maximum assessment limited to $5,000/day. Payment: Monthly, lOt retainage. Construction method: Soft-ground shield with breasting facilities. Restrictions: Hauling according to applicable county ordinances. Noise levels for equipment in various locations and hours of res- ident activities. Disputes resolution: Decision by Owner's contracting officer. De- cision can be appealed within 30 days to owner's board of direc- tors~ board decision final unless question is one of law. Geotechnical data aade part of contract docuaentaa Boring logs (bound directly into the contract drawings). Core samples, spe- cifically indicated as available for inspection on 24 hours notice. Geotechnical reports, with profiles and results of all field and laboratory testing on soil samples, boreholes, and observation wells. (The geotechnical reports were laid out for bidders• exami- nation and copies could be obtained from the National Technical Information Service.) Disclaimers& Apparently none with respect to owner-furnished in- formation on subsurface conditions. Changed-conditions clause& Yes Construction Method& Soft-ground shields with excavator hoe, breast- ing doors, and articulation capabilities. Ptimary support of ribs and lagging. Permanent support of cast-in-place concrete. Conditions Encountered& Water inflow of up to SO gpm from clean sand lenses. Hard, cemented sand lenses and layers up to 4 ft thick for 1,000 ft. Alternating pervious/impervious layers. Wet, flowing single- size sand lenses for 1,200 ft of each tunnel. Unstable ground around existing sewer. Problems Encountered& Construction& Hard sandstone lenses and layers required instal- lation of rock-breaking hoe rams in shields and resulted in very slow advance rates when the lenses and layers were encountered. Single-size flowing sand lenses were difficult to dewater because of intervening clay layers and resulted in runs, major voids and settlements, and bogging down of both shields so that progress averaged only 30 ft/week. Wet, sandy ground around existing sewer 60 

required grouting for stabilization, with the result that the shield was practically grouted in place and the hood buckled when shoving resumed. Operationa and Maintenance• Acid water resulted when the shields penetrated about 2,000 ft of ground rich in iron sulfide (FeS 2 ), which oxidized when exposed to air in the advancing dewatered tun- nels and formed sulfuric acid. The pH values of the groundwater in the affected area ranged as low as 2.0, which raised concern about corrosive effects on the permanent concrete lining. Initial studies indicated that the outside of the tunnels reacted with the acid and associated sulfate ions to create an impervious layer which effectively blocked further attack. Additional studies are being pursued to determine if this holds true and if the acid may be dissipating with time. The permanent lining in the vicinity of the acid water problem was extremely leaky after completion of construction. Effects of water intrusion were made worse by masses of muddy, rust-colored ferrous and ferric hydroxide, Fe(OH) 3 [a by-product of the acid formation] which formed troublesome deposits on walls, inverts, and safety walks. Three overlapping programs of post-construction chemical grouting were necessary to dry up this stretch of tunnel and prevent the rusty intrusions. Resolution of Assertions Re Subsurface Changeaa The contractor as- serted that hard sandstone lenses and layers had been encountered where only soil had been expected. The owner agreed that the hard sandstone was unexpected and paid the contractor extras as the lenses and layers were encountered during construction. The total payment was $940,848. The contractor asserted that neither (a) the combination of single size flowing sand lenses and intervening clay layers, nor (b) the un- stable ground around the sewer were expected to be encountered. The owner disagreed, maintaining that these conditions could be easily pre- dicted from information in the contract documents. Litigation ensued, with the contractor claiming +$22 million, about 95 percent of which pertained to (a). Litigation before the Corps of Engineers Board of Con- tract Appeals proceeded through the pleading, discovery, and trial phases; the parties achieved settlement on their own before a final de- cision by the Board. It may be significant that the figure settled on after the start of litigation--$3, 777,463 for claims (a) and (b) --was part of a three-contract closeout settlement in which the contractor re- covered $7 million out of claims totaling $SO million. Analysis/Opinion• The primary fact about WMATA's site investigations and their impact on construction is that when this project was let for bid, in 1975, the •subsurface investigation reports• (WMATA's term) were not made a part of the contract documents. The boring logs, which are presumably mostly factual, were bound into the contract drawings and the bidders were responsible for the information contained therein. However, the subsurface investigation reports contained much interpretive data for which the owner did not wish to be held completely responsible. The reports were made available for study during the bidding period and 61 

photocopies could be purchased from the National Technical Information Service, however, this really served to confirm their status as informa- tion docWilents rather than binding contract docWilents. The situation may have made it justifiable for the G-2 contractor to make the somewhat ambiguous statement that the reports were •studied, read, and respected, but could not be relied upon.• The legitimacy of this argument might be disagreed, but it is difficult to dispute because the contractor was later able to assert successfully that he was not required to take into account the available geotechnical information in making his bid. The owner's arms-length attitude about his own reports may have worked to his disadvantage because the reports seemed to document very nicely the wet, flowing sand conditions that turned out to be the great- est problem on the job. Although no pWilping tests were performed, there was good borehole coverage, plenty of falling head tests, and more than enough lab testing to define the nature of the soils adequately. Al- though one might quarrel with the lack of pWilping tests, the fact is that the owner's geotechnical engineer was able to use the available data to describe the wet, single-size sand lenses that would be diffi- cult to dewater because of the intervening clay layers. Had the con- tractor relied on that information--which he might have done if the sub- surface investigation reports were considered full-fledged contract documents--he might have based his bid on more stringent dewatering and/ or a better breasting system, thereby avoiding some very costly delays from bogged down shields. The cost of a more conservative original de- watering plan and better breasting equipment would not have approached the $21 million (plus or minus) claimed for flowing ground, and would have been much less than the $3.6 million (plus or minus) finally set- tled on that claim. In addition, if the contractor had heeded the predictions of flow- ing ground and not suffered such slow progress for 1,200 lin ft in each tunnel, a secondary problem might well have been avoided. The sulfuric acid that materialized in the ground where the shields were struggling was apparently caused by oxidation of minute crystals of pyrites (i.e., iron sulfide, FeS 2 ). This probably would not have occurred had the machines made normal progress, but the day-after-day exposure of soils in an aereated tunnel face subjected them to an oxidizing environment that normally would not be encountered except in an excavation complete- ly open to the surface. The resulting acid condition was not a severe excavation problem and caused no claim, but the oozing by-products made it mandatory to provide for extra-thorough grout sealing of the com- pleted tunnel, while at the same time interfering with the grout's ef- fectiveness and causing it to be more expensive. The acid also caused several years of additional study expense and general unease before it was concluded that the acid would not harm the concrete lining and would dissipate before causing any environmental damage. WMATA has never been faulted for failing to recognize the acid-producing potential of the ground, because it is a rare condition heretofore known only to a few very specialized soil scientists who have documented the behavior of the so-called •cat clays • in road cuts and other excavations open to the surface. It is worth noting that the condition apparently has not oc- curred in other WMATA tunnels, but will probably have to be watched for in the future. 62 

The other major subsurface-related construction problem lay in the hard, cemented sand lenses whose identity and difficulty of excavation were not suspected by the owner or his geotechnical engineer. Actually rock-like in consistency and up to several feet thick, these lenses were either penetrated by a tri-cone bit or punched through with a split spoon during the investigations. No samples were recovered, and the re- sulting high blow counts were assumed to indicate only the presence of cobbles or boulders. The contractor obviously made the same assessment because he was surprised when his shields kept hanging up on what were in essence small masses of sandstone. By quickly negotiating extras worth nearly $1 million, the owner admitted that his site investigation was deficient in its techniques of identifying hard materials. Inter- estingly, however, it is very possible that pre-bid identification of the cemented sands would have made little difference in the ultimate cost of the tunnel. If the contractor had been fully aware of the difficulty of excavation, he might theoretically have increased his bid price by about the same amount as was ultimately negotiated in the field anyway. And, of course, the owner kept the problem from escalating beyond its true value by admitting fault and negotiating rather than entering into expensive litigation. As of this writing, the washington Metro system is about 22 years beyond the date of its first feasibility site investigation, yet final design investigations for some sections are currently under way and many others are scheduled for the future. This creates an opportunity for the owner to use construction feedback to vary his site investigation philosophy in order to respond to newly perceived conditions, to learn from experience, and to not continue with faulty methods. The Section G-2 case history provides the following examples of how WMATA has insti- tuted such changes: • In the sampling of certain sedimentary materials, tr i-cone bits and split spoons are withdrawn from the borehole at the first indi- cation of hard drilling that might signal the presence of a cemented sand layer. Then a diamond bit core barrel is substituted and the en- tire thickness of the layer is recovered for proper logging and ultimate examination by bidders. • When drilling in potentially acid-producing ground, WMATA geologists now watch carefully for the presence of fine crystals of sul- fidic minerals (generally pyrite and marcasite), especially in dark colored soils that might indicate deposition under reducing conditions. A few soil samples from each drilling program are routinely lab-tested for •total sulfur• content to detect the presence of sulfide concentra- tions that might escape detection by eye. Any time there is doubt about a soil's acid-producing potential, a consulting specialist is called in to render an opinion on the subject. Possibly more important than the above changes in site investiga- tion techniques is WMATA's relatively recent decision to upgrade the subsurface investigation reports from their status of information docu- ments to full-fledged contract documents. Of course, the change came about because of many episodes of litigation on many projects, but the G-2 case history of flowing sand is a perfect example of why such a 63 

change may have been needed. It will no longer be possible for a con- tractor to deny responsibility for knowing the contents of the reporta1 rather, it will be expected that the bid and construction planning are baaed on that knowledge. The decision to upgrade the statue of the re- ports may make the owner more certainly liable for mistakes in interpre- tive information. However, it should also create much more consistency in the assumptions made by bidders and · will definitely curtail much time- consuming argwaent over whether a contractor ia to rely on all of the information provided. 64 

CASB S'rUDY RO. 4 aa.a of Projecta Bonneville 2nd Powerhouse Railroad Tunnel Purposea Railroad tunnel relocation Locationa Columbia River Basin, Washington and Oregon (42 miles east of Portland) Construction Perioda June 1976--September 1977 Site Investigation Perioda November 1974-~arch 1976 Siaea 1,338 ft longJ 35.9 ft high by 24.3 ft wide. Project Costa Estimated $8,636,558 Bid $10,410,610 As Completed $12,172,226 (includes all extra payments) Mined Tunnel Construction Costa Estimated $5,834,261 (excluding profit) Bid $7,246,650 Extra Support Contract Mods $1,279,674 Subsurface Related OVerruns $0 As Completed $8,526,324 Subeurface Investigation Costa $1,452,026 pre-bid (excluding profes- sional services). Sum.ary of Site Geologya Unconsolidated cascade landslide deposits consisting of igneous, pyroclastic, and sedimentary slide debris. A graded mixture of gravelly, silty sand surrounding some basalt boulders and slide blocks of Wiegle formation sandstone/siltstone/claystone/ conglomerate/lava with bedding dips of no more than 25 degrees and occa- sional high angle shear zones. · Blocks soft to moderately hard with un- confined compressive strengths of perhaps l to 3 ksi, generally weakened by their movements. Extremely variable materials defined in the con- tract as mixed face. Depth of overburden ranges from 28 to 190 ft sur- face to crown. The mass contains highly variable percolating water, perched water tables, trapped water, and flowing zones. Rainfall re- charges these areas, and the primary aquifer located in a layer of allu- vium well below tunnel invert is hydraulically connected to the Columbia River. Minimal tunnel inflow expected, however, because the groundwater table is below invert most of the year. Design Criteriaa Assumed vertical rock load of 35 ft (one tunnel height) for temporary support. Assumed water levels would be drained to below invert level before or during construction. 65 

Contract Proviaionaa Typea Unit price per cubic yd of excavation and unit prices for support items (steel sets, rock bolts, shotcrete). Estimated quan- tity variation limits not specified. Stipulations a Schedule and/or tiae of completion. Liquidated damages: $4,285 per calendar day of delay. Payment: Monthly; lOt retainage (may be reduced after SOt comple- tion). Progress payments for support items. Construction method: Not specified but subject to approval of con- tracting officer. Restrictions: Blasting (minor). Disputes resolution: Decision by owner's contracting officer con- cerning questions of fact arising under the contract. Only re- course was litigation. Geotechnical data -de part of contract docu.entaa Detailed geo- technical report describing conditions in the pilot tunnel. Geolog- ic profiles provided in the drawings; material classification maps provided in pilot tunnel section of contract. Mechanical analysis (gradation curves) provided in drawings of pilot tunnel samples. Diaclabaeraa General Provision 141 stated contractor is respon- sible for estimating properly the cost and difficulty and that the government assumes no responsibility for available information. Changed-conditions clauaea Yes Construction Methoch Top heading and bench and drill-and-blast, with drilling jumbo, wheel muckers, rebar jumbo, and lining form. Primary support of steel sets, rock bolts, shotcrete, and concrete wall plate. Permanent support of cast-in-place reinforced concrete (21 in.) and mis- cellaneous steel. Conditions Bncountereda Essentially as predicted by owner information. Problems Bncountereda Construction• Minor fault problems, squeezing and running ground. In some areas there was inward movement of high side walls, contained with tiebacks and invert struts. Some water in- flow, but effectively controlled by dewatering and grouting (as- sisted by favorable drought conditions during the construction period). Operations and Maintenance• None of any consequence identified. Reaolution of Asaertions Re Subsurface Changeaa No assertions made. 66 

Analyaia/Opiniona This tunnel is a good exaaple of the effectiveness of a thorough and well-conceived geologic site investigation in keeping the costs of tunnel construction down. The site was known to be very risky, so this short (1,400 ft, plus or minus) tunnel had a geophysical study, thorough surface mapping, 54 boreholes of various types, and a pilot tunnel with a geologic report on the pilot tunnel. With this in- formation, the contractor was prepared for any conditions and was able to complete on time and with no claims for differing site conditions. In this case, the site investigation cost was about 12 percent of the bid price, but without these investigations the bid certainly would have been higher and the final cost would have been much higher. 67 

CASB STUDY HO. 5 aa.e of Project1 Buckskin Mountains Tunnel (Spec. No. DC-7096) Purpose1 Water conveyance Location1 Arizona (20 miles northeast of Parker) Construction Perioda April 1976-~ay 1979 Site Investigation Perioda 1967--1972 Size1 35,915 ft long; 23 ft 5 in. diameter. Project Coati Estimated $53,804,499 Bid $58,256,638 As Completed $65,613,963 Mined Tunnel Construction Costa Estimated $49,627,190 Bid $47,268,690 General Contract Mods $1,000,367 Subsurface Related overruns $5,441,077 As Completed $53,710,134 Subsurface Investigation Coat1 $1,238,000 estimated pre-bid SUIIIIIary of Site Geologya Volcanic flows in mass landscape of the Buckskin Mountains are dominated by andesite interlayered with tuff and agglomerate, which have been intruded by andesite dikes and laccoliths. The andesite is hard, dense, and blocky; it is situated in rather flat- lying flows and ranges from 10 to 100 ft thick. Pyroclastic rock inter- flows are 5 to SO ft · thick. The andesite exhibited few weathered zones and is quite strong, with unconfined compressive strengths up to 43,500 psi. The tuff and agglomera~e, poorly to well indurated, is cemented with gypsum and calcite; the unconfined compressive strength is 1,100 psi. One fault zone was identified on the surface near the outlet portal. Design Criteria• Maximum rock load of 70 ft; hydrostatic head ranges from well below tunnel invert to 18 ft above crown. Contract Provisions• Type a Unit price per linear ft of tunnel excavation and unit prices for precast liner segments (per square ft) and installation (per linear ft) • Stipulation& I Schedule and/or time of completion: 30 months for tunnel excavation and support; 1,800 calendar days for the total contract. Definition of delay and suspension of work. 68 

Liquidated damages: $2,000 per day. Payment: Monthly; lOt retainage (up to SOt completion). Construction method: Contractor's option. Restrictions: Environmental precautions caused by mating of Blue Heron. Disputes resolution: Initial decision by contracting officer, with appeal possible to head of agency. Further appeal to Board of Contract Appeals, Department of the Interior. Geotechnical data ude part of contract docu.ntaa Preconstruc- tion geologic report; surface geologic map, profile, and boring logs included in contract drawings. Gravity survey results available for inspection. Construction and foundation materials teat report available as separate document, by request only. Corea available for inspection; samples (up to 30 in.) permitted for testing. Diaclaiaeraa Borings show conditions at locations drilled only. Any interpretations are strictly the contractor's responsibility. Changed-condition& clauaea Yes Construction Metboch Tunnel boring machine with flexible, articulated hood, side grippers, and 15-1/2 in. disc cutters. Permanent (and pri- mary) support of 6- to 7-in. thick segmented rings (reinforced, precast concrete). condition& Bncountereda Loose joint systems and blocky ground condi- tions in the hard andesite. Soft invert in the tuffs and agglomerates. Fault zones with running ground conditione and cave-ins. Construction• Loose joint systems in the andesite resulted in blocky rock, face fallout, and excessive overbreak conditions which obstructed the mucking system and damaged the cutterhead compo- nents, requiring coaaplet-e redesign of the cutterhead. Widely spaced joint systems resulted in blocky rock and roof fallout con- ditions which greatly reduced and sometimes stalled 'l'BM progress. Soft rock in the invert caused the TBM to dive and resulted in problems with alignment and grade. Fault zones resulted in a cave- in and raveling ground conditione (chimney) at two locations. Face fallout at two locations required grouting and concreting of cavi- ties ahead of the TBM. Operation& and Maintenance• None Resolution of Assertion& Re Subsurface Oaangesa The contractor filed claims totaling $7,767,802 for the above cited construction problems en- countered. The claims were denied by the owner for the llOSt part, and then settled by the Department of Interior Board of Contract Appeals. The Board awarded the contractor $5,441,077, but $1,343,077 of that amount was interest on the settlement award of $4,098,000. The method 69 

of calculating the interest is a subject of dispute, still unresolved as of the interview date. Analyaia/Opiniona The Buckskin Mountains Tunnel was bid in January 1975. Contractors had three options. Schedule No. 1 called for drill- and-blast excavation with a horseshoe shaped tunnel using cast-in-place final lining. Schedule No. 2 was for aachine boring caabined with cast- in-place final lining. Schedule No. 3 was for machine boring and pre- cast concrete liners. (Schedule No. 4 was for non-tunnel items.) The low bid was on schedule No 3 and represented the first attempt in North America to use four-piece precast concrete rings. The contractor ordered a special hard-rock tunnel boring machine. This TBM was designed to cut 40,000 psi andesite and supported a 360- degree shield that served as a form for the four 3-1/2 ton precast lin- ing segments. Each tongue-and-grooved ring comprised 5 ft of the tunnel length. One hoist placed the invert segment on a bed of preshaped pea gravel. The side and crown segments were carried into position by a ring gear located inside the inner circumference of the tail shield. Once the segment was rotated to the proper elevation and the tongues and grooves aligned with the previously placed sections, rams were used to push the segment into place. Lining was then completed by sealing the joints with mastic, blowing pea gravel into the annular space outside the segments, and grouting the gravel. The invert waa grouted twice a week and the rest of the ring was grouted at longer intervale. The TBM was advanced by a rib gripper system and the lining waa not designed to react against the machine's forward thrust. The preconstruction geotechnical investigation waa felt to be thor- ough by both the contractor and tunnel owner. There was no pre-bid con- ference. All interested parties were encouraged to viait the tunnel site and view the rock cores. Up to 30 in. of core was made available to any plan holder wishing to conduct his own tests. Two major geotechnical problems were encountered during construc- tion. Small cave-ins tended to chimney upward placing heavy rock-weights on the shields. Second, at the face, the 15-1/2 in. diameter cutter discs projecting from the cutterhead along with the muck buckets (or scoops) tended to catch hold of the rock blocks and pluck them from the tunnel face before they could be broken into small enough pieces to be carried away by the muck handling system. In general, the problem only occurred with rock pieces larger than 6-in. cubes. These rock pieces resulted in considerable damage to the cutterhead components and mostly to the muck scoops. It is quite possible that a mechanical rock core log (i.e., discontinuity spacing determination, piece counts, etc.) of the drill holes could have provided a forewarning of thia loose joint problem. TO relieve the problem with the rock pieces, a false face was built on the cutterhead. Thia reduced the projection of the 15-1/2 in. diame- ter cutters to 4 in. In addition, low profile muck scoops were placed on the cutterhead circumference. The apace between the original cutter- head and the false face was filled with grout to provide mass for vibra- tion dampening. TO withstand the cave-ins and keep large rock fallout from binding the machine, the shield was changed from the original l-in. thick plate to 1-1/2 in. thick plate and extra internal bracing was 70 

added. Modification of the machine required more than three months to complete. Another problem was a soft tunnel invert which caused the TBM to sink below grade. Gypsum had been identified in the preconstruction site investigation1 however, water was not considered a problem (average annual rainfall less than 10 in.) and the core holes were not backfilled with grout as practice should normally dictate. It could well be that the drill holes open to the surface helped precipitate some of the soft invert problems. The preconstruction geologic study missed locating two fault zones encountered during excavation. 'l'he boring logs gave no indication of open joints or blocky rock conditions. Drill holes were generally spaced 3,000 ft apart on this 35,915 ft tunnel (only 500 ft apart near the portals). All told, nearly six months of construction time were lost due . to unforeseen geologic probleiiS that caused a major rebuild of the TBM. The major claim was for the TBM rebuild to accommodate the loose, blocky and raveling ground conditions. The ground conditions also affected the lining. The precast seg- ment design was based on a 70 ft rock load and a maximum deflection of 0.5 percent of diameter. The concentration of blocky rock loads failed several rings and the ring sections were redesigned while the TBM was being overhauled. Several important lessons vital to tunneling in potentially blocky rock were apparent on this project: • Geologic problems can be interdependent with the selected ex- cavation method and must be considered as a necessary part of the pre- construction site study. With such a study, the blocky rock conditions would not have caused the three-month delay for rebuilding the TBM. • Detailed knowledge of the joint spacing, openness, roughness, and filling is necessary in any rock formation that could have blocky rock. use of a mechanical drill core log and angled drill core perhaps could have relieved some of the problems. • Even when water is not expected to be a problem, all core holes should be grouted bottom to top to prevent ingress of surface water. • A properly conducted water make/loss study may have helped to identify the loose joint system. In all this, the tunnel proved to be an outstanding demonstration of a good mining method. Once the TBM was refurbished, the average 24- hour production day was 70 ft of excavated and final lined tunnel. 7l 

CASB S'rUDY HO. 6 .... of Projecta Hades and Rhodes Tunnels (Spec. No. DC-7421) Purposea Water conveyance Locationa Utah (40 miles northwest of Duchesne) Construction Perioda September 1980--November 1981 Site Investigation Perioda June 1975--Summer 1978 Siaea 26,259 ft long (Hades • 22,149 ft, Rhodes • 4,110 ft); 10 ft 5 in. diameter. Project Costa Estimated $35,494,430 Bid $34,681,703 As Completed $34,611,894 estimated Mined Tunnel Construction Costa Estimated $32,951,695 Bid $27,908,413 Subsurface Related Underruns $1,737,425 Subsurface Related overruns $1,380,086 As Completed $27,551,074 Subsurface Investigation eo.ta Not available Suaary of Site Geologya Alternating strata of limestone, sandstone, siltstone, and shale, dipping at 18 degrees in a regional hOIDOCline. Bedding ranging from thin to thick, with some of the lilllestone being massive. Closely to widely spaced joints. OVerall quality varying widely: shales generally weak, sometimes squeezing and swelling; lime- stone often solutioned; sandstone generally hard and sound, but cementa- tion somewhat variable. At least three faults identified. MaximWil overburden (surface to crown) 2,200 ft at Hades and 590 ft at Rhodes. Except for the black shale, all strata water bearing and expected to produce tunnel inflows of 1,000 gpm at Rhodes (diminishing with time) and 3,000 to 4,000 gpm at Hades (for extended period) along with exten- sive groundwater reservoirs. Design Criteriaa Head ranges from -40 to +200 ft above crown according to borings. Range of ground loads not available. Contract Provisions• Typea Unit price per linear ft of finished tunnel, except for pressure grouting. A second unit price was requested for quanti- ties beyond a specified limit. 72 

Stipulations a Schedule and/or time of completion: 1,445 calendar days for total contract. Definition of delay and suspension of work. Liquidated damages: $2,000 per day. Payment: Monthly, lOt retainage (may be reduced after 50\ comple- tion). Construction method: Options were (1) drill-and-blast for horse- shoe, (2) drill-and-blast for circular horseshoe, (3) TBM for circular, and (4) •tunnel excavating machine• for modified horse- shoe with vertical sidewalls. Restrictions: None indicated. Disputes resolution: Decision by contracting officer. Subject to appeal to head of governmental agency whose decision is final un- less question is one of law. Geotechnical data aade part of contract docuaentaa Summary of the geological investigations from specifications; draft of precon- struction geologic memorandum available for inspection. Surface geology map and diagrammatic geologic sections for each tunnel. Photos of core samples in contract documents and cores available for inspection. Electrical resistivity logs and results of ex- pansion and uplift tests on shale samples. Disclaimers& Deductions, interpretations, and conclusions from factual information are sole responsibility of the contractor. Changed-condi tiona clause a Yes Construction Method& Tunnel boring machine with two grippers and 14- and 12-in. cutters. Primary support of steel ribs and rock bolts. Per- manent support of unreinforced cast-in-place concrete (16 in.). Conditions Encountered& The rock types were as predicted, with much of the shale exhibiting definite squeezing tendencies but with the solu- tioned limestone producing a much greater volume of water than expected. Poor ground stability in mud filled cavities in limestone. Running ground in sand for a 50-ft reach. Problems Bncountereda Construction& Excessive overbreak and squeezing shales appear to have affected operations, but not sufficiently to drive the contrac- tor's costs above the figure that was bid. The most serious problem lay in five areas of Hades where large quantities of groundwater flowed from mud filled solution cavities in the limestone. Total flow reached as high as 6,000 to 8,000 gpm for extended periods, at least twice the quantities predicted by bid documents. The occur- rences resulted in a number of delays to allow the flows to dissi- pate and made necessary an upgrading of the pumping system and periods of hand mining. 73 

Operations and Maintenancea No problems identified. Resolution of Assertions Re Subsurface Olangeaa The contractor made a claim (amount unknown) for the excessive water pouring from the Hades solution cavities. While the work was in progress, the contracting offi- cer acknowledged the claim's validity and negotiated extras worth $62.31 per lin ft of the entire tunnel. The changes to construction involved deletion of the pressure grouting intended to plug the cavities and pay- ment of $1,380,086 ($62.31 x 22,149 lin ft) to deal with the large water inflows. Because the cost of the grouting would have been greater than the cost of water handling, this change resulted in a net reduction in the contract price for the tunnel itself. Analysis/Opiniona The general nature of most potential problems could be estimated ahead of time, based on experience in similar geologic set- tings. However, the owner did not provide sufficient information to permit bidders to make reliable quantitative estimates of problems (e.g., water, squeeze). Hence, the bidders were forced to take a great element of risk. Due to the great depth of overburden, the ability to explore thoroughly with borings was severely limited. The owner did a reason- able job of defining general stratigraphy based on published literature but did not provide cores for the entire stratigraphic section. Also, the owner did not provide sufficient data from laboratory tests to en- able reliable estimates of squeeze behavior of weak shales; bidders had to estimate behavior based on their previous experience. Detailed in- formation on experience with two nearby tunnels (different formations but similar overall geologic setting) was available to the owner but not provided to bidders. This project was successful, but not because of excellent and ade- quate geologic and geotechnical data. Rather, its success can be at- tributed to the following: • The contractor developed efficient means of handling diffi- cult ground conditions such as heavy water inflows, squeeze, and exten- sive overbreak/fallout. • Some problems were less severe than possible (e.g., geologic conditions were present for potentially even greater water inflows). • The owner was willing to negotiate changes with the con- tractor. 74 

CASB ftUDY NO. 7 ..._ of Project1 carley v. Porter Tunnel (No. 65-29) Purpoee1 Water conveyance Locationa california (Kern and Los Angeles Counties) Construction Period1 April 1966--october 1969 Site Investigation Perioda 1957--1965 (very intermittent) Siae1 25,075 ft long; 24 ft 4 in. diameter. Project CO.ta Estimated $42,321,830 Bid $33,788,800 As Completed $48,316,215 Mined Tunnel Construction Coati Estimated $41,341,900 Bid $32,848,600 Extra Support Contract Mods $11,369,256 Subsurface Related Claims $2,500,000 As Completed $46,717,856 Subeurface Investigation Coati $2,000,000 estimated pre-bid S.-ary of Site Geology1 Mostly highly fractured, locally altered, strongly crushed and sheared Tejon Lookout granite with roof pendants of metalimestone and hornfels. Rock quality extremely variable, but gen- erally poor. Garlock fault crossed inlet portal; many subsidiary shears throughout alignment. Some lakebed deposits consisting of siltstone and claystone, poorly to moderately indurated. Heavy, locally squeezing and/or running ground expected. Water stored in fractures, shear and granular zones, generally occurring in sporadic pockets. Depth of over- burden ranges from 0 to 1,800 f.t surface to crown. Design Criterial Maximum ground load of 18,150 psf (calculated from load cell data in pilot tunnel); up to 1,520 ft head of water above crown. Contract Provisional Type1 Unit price per cubic yd of excavation and unit price (by weight and quantities used) for temporary and final lining compo- nents. Estimated quantity variation limits not specified. Stipulation& I Schedule and/or time of completion: 1,330 calendar days for total contract. Definition of delay and suspension of work. Liquidated damages: $2,625 per day. 75 

Payment: Monthly, lOt retainage (may be reduced after SOt comple- tion). Disputes resolution: Decision by owner's engineer. Decision can be appealed initially to division chief and then to contract appeals board. Geotechnical data made part of contract cJocUJMntaa Geologic data report available on request, including a profile with stick logs, geologic maps of route and pilot tunnel. Geophysical logs, core samples, and test data available for inspection. Disclaimers& Data provided for information only. Conclusions and interpretations are sole responsibility of the bidders. Changed-conditions clausea Yes Construction Methoda Two hydraulic shields with pushing jacks and forepoling jacks; drill-and-blast used as necessary. Primary support of steel liner plates with steel sets and gunite as needed. Permanent sup- port of unreinforced cast-in-place concrete (10 in. minimum). Conditions Bncountereda Highly fractured, locally sheared, altered and crushed Tejon Lookout granite with roof pendants of meta-limestone and hornfels. At outlet portal, soft Pliocene lake deposits consisting of flat bedded, poorly to moderately indurated siltstone and claystone. Local heavy, squeezing and running ground. Water apparently stored in fractures, shears and granular zones, occurring mostly in sporadic pock- ets. Conditions quite variable, but generally very poor due to crossing the major Garlock fault and its many subsidiary shears. Depth of over- burden ranges from 0 to ±1,800 ft surface to crown. Probleaa Bncountereda Construction& Running ground and blocky ground, which in one in- stance caused a tunnel collapse that trapped 17 men for 22 hours and caused a 5-month delay for remining. One zone of very high water pressure and squeezing ground caused pressure binding and structural collapse of the shield. There was overall difficulty in steering the shield so that it failed to maintain the specified alignment. There was general slow progress at 29 locations where faults, granu- lar and clayey altered granitic materials, large water inflows, running ground, and heavy ground loads were encountered. OVerall conditions were so difficult that the contractor mobilized a shield and substituted steel rib and steel liner plate for the steel rib and rock bolt initial support called for in the contract documents. Operations and Maintenance& One low area caused by diving of the shield was found to be silting up when the tunnel was inspected about 10 years after completion. Resolution of Assertions Re Subsurface Changesa The contractor filed two major claims, one for the collapsed, pressure bound shield (zone of 76 

high pressure water and running ground) and another for the generally slow progress at 29 locations, ~~aintaining that the conditions were unusual, unknown, and 110re frequent and severe than anticipated. The two claims totaled $7,870,101. The first claim was denied completely by the owner, stating that shield collapse was caused by contractor procedures and poor condition of the shield. The owner also filed a $300,000 counterclaim for failure to Mintain the specified alignaent as well as negotiating a support steel unit price reduction worth $2,626,300 to himself. The owner disagreed with the contractor's second claim, indicating that the condi- tions were known, less severe than predicted, and that contractor proce- dures had contributed to the proble... However, the contractor's second claim was settled at closeout (without litigation) for $2,500,000, which was 32 percent of the amount requested. Analysis/Opiniona This project is a prime example of how difficult it can be to clarify the question of subsurface related cost overruns in judging the adequacy of the pre-bid site investigation. The simple tabulations indicate that the contractor asked for extras totaling $7,870,101 to cover the cost of a shield collapse and the encounter with unexpected difficult tunneling conditions at 29 locations. The owner negotiated extra payments of $2,500,000 for the difficult tunneling and the records shew that amount of loss due to claims. However, this pic- ture may be oversimplified because the tabulation also shows that an additional $11,369,256 in contract modifications was paid to cover the cost of extra tunnel support, moat of which was steel used in continuous liner plate proposed by the contractor and approved (though hardly an- ticipated) by the owner. Because the added support seems attributable to geologic conditions, it is probably fair to say that true overruns really amount to more than 5-l/2 times the 7.6 percent that would be indicated by looking at the claims alone. The subject tunnel traversed an extent of ground that can be de- scribed as extraordinarily bad, given the excavation and support methods available. The inlet portal was driven through part of the major Gar- lock faultJ the entire tunnel was driven through a wedge of ground caught between the Garlock and the San Andreas faults so that it was ex- tremely fractured, sheared, crushed, and altered. The general condition is sUllied up in the statement from the •as-built• geology report that faults were 11apped at an average spacing of ll ft along the entire tun- nel length. The seriousness of the condition was highlighted early by the low bidder's opting for soft-ground shields and continuous steel liner plate in what was supposed to be a rock tunnel that could presum- ably be supported initially with steel ribs and rock bolts, according to the contract documents. Although the alignment was apparently set in the best available location after extensive study of alternatives, it was moat certainly a case of choosing the lesser of known evils. This ground obviously deserved the 1108t thorough of site investiga- tions. The owner approached it by relying mainly on a program compris- ing boreholes, a test adit, and a pilot tunnel. The 600-ft test adit and the 3,688-ft pilot tunnel cost well in excess of $1,350,000, more than four percent of the engineer's estimate of tunnel construction cost. They did help to indicate the frequency of poor rock zones that might be 77 

encountered and provided the opportunity for grouting program evaluation, water inflow observations, overbreak monitoring, and ground load cell installations. The 1,250-ft spacing for exploratory boreholes seeaas a bit wide considering the poor geology. The choice of spacing was most likely influenced by difficult access and some great penetration depths (up to 1,820 ft), factors which could have made the desirable number of borings cost-ineffective. In such variable geology, with so many zones of poor rock, it would require an extremely close borehole spacing to thoroughly delineate the ground conditions for the entire length of tun- nel. It may not have been unreasonable for the owner to put more than two-thirds of his exploration budget into a test adit and pilot tunnel for thorough observation of the ground at least along those limited lengths. This statement assumes that the owner should then be able to extrapolate from those limited area conditions to judge the general con- dition of the main tunnel extent. The necessary extrapolation may not have been accurately carried through, but the responsible parties must at least be credited with a serious overall effort. Indications are that the owner did a considerable amount of de- tailed geologic interpretation before deciding that Carley V. Porter could be initially supported in typical rock tunnel fashion, with steel ribs and rock bolts. The contracting system was flexible enough to com- pensate for a certain degree of underestimating because support iteaas were unit priced and it would be easy to pay the contractor for adding jump sets, additional bolts, etc. Yet the owner must have realized early that his assumptions had been very optimistic because he quickly accepted the low bidder's plan to drag soft-ground shields through what should have been rock, supporting it with a system normally associated with earth. He may have been very surprised later by the ultimate proj- ect cost because much of the money above and beyond the bid price went for an extensive number of unit priced steel ribs in addition to the agreed upon heavy steel liner plate. Nevertheless, the quick admission that previously presumed hard ground deserved some soft-ground treatment indicates that the owner had doubts about either his subsurface data or his interpretation of them. Regardless of whatever doubts the owner may have had about his own geologic interpretations, . he should have disclosed them to the bidders. This was not done, however. The contract geologic data consisted mostly of nondetailed stick boring logs, with no geologic profiles and almost no interpretive information. Working with this limited body of knowl- edge, the low bidder was apparently able to perceive the nature of the ground, at least in a general way, more accurately than the owner. Had the bidder been able to examine the detailed, interpretive information with a construction attuned eye, even more of the problem would possibly have been apparent earlier in the ga., which then might have led to better overall planning. The owner may have sought protection by with- holding pertinent knowledge in the belief that he then could not be held strictly accountable for possible misinformation. However, the bottom line is that the project overruns directly attributable to ground condi- tions amounted to 33.5 percent of the engineer's estimate and to 42.2 percent of the low bid. The owner gained little from the restrictive disclosure policy and may have actually lost money by employing it. 78 

CMB ftUDY NO. 8 aa.e of Projecta Red Book Interceptor Sewer (Contract lA) Purpoaea Sewage conveyance Locationa New York (Brooklyn) Conatruction Perioda April 1978-~ay 1980 Site Investigation Perioda Early aid 1969--June 1970 Siaea 8,600 ft long, 10 ft 5 in. diameter. Project eo.ta Estimated $55,733,229 Bid $61,862,009 Mined Tunnel Construction Costa Estimated $50,242,060 Bid $52,283,285 General Contract Mods $ -80,168 Subsurface Related OVerruns $935,999 As COmpleted $53,139,116 Site Investigation eo.ta $74,000 pre-bid (post award costs not avail- able). S.-ary of Site Geologya Mostly granular, miscellaneous fill overly- ing clean, horizontally bedded fine to medium glacial outwash sands with soae gravel. Frequent channels, pockets, and lenses of bouldery till and peaty clay. Some obstructions expected in the form of timber piles, bulkheads, and piers. Sands generally compact to very compact, but some loose spots. Permeability of sands and amount of available water great enough to require compressed air or slurry shield for control. Depth of overburden ranges from 12 to 70 ft surface to crown. Deaign Criteriaa 2 to 10 ft head of water above crown. Contract Provisionaa Typea unit price per linear ft of completed tunnel and unit prices for grout and removal of boulders. Estimated quantity vari- ation limits not specified. Stipulations a Schedule and/or time of completions 1,100 calendar days for total contract. Definition of delay and suspension of work. Liquidated damagess $2,000 per calendar day. Payment: Monthly, 7.5t retainage (until 5t of total contract retained). Construction methods Slurry shield or shield with compressed air. 79 

Reatrictionss Blasting subject to engineer's approval. Dewatering not allowed near anchorage of Brooklyn Bridge' elsewhere it was limited to a maximum 6-ft lowering of the water table. Disputes resolution: Decision, which is •final,• by Commissioner of Water Resources. Appeal possible to commissioner, only other re- course is litigation in court. Geotechnical data aade part of contract docuaentsa Report of soil investigation for proposed tunnel section, dated June 1970. General profile included in soil investigation report. Diaclaiaers a Data furnished for information only and not a sub- stitute for personal investigation. Changed-conditions clauaea Yea Construction Metboch Pull breasting, soft-ground shield with hydrau- lic excavator and using 18 psi compressed air. Primary support of heavy steel liner plate. Permanent support of cast-in-place concrete. Conditions Bncountereda Essentially as predicted by owner informa- tion, except that boulders and timber obstructions were far more numer- ous. (The selected contractor suspected this possibility during the bidding period because he performed his own subsurface investigation.) In addition, natural gas (methane) and man-made toxic wastes were en- countered. Probleu Bncountereda Constructiona Running ground was severe enough to require full breasting in spite of the compressed-air operation. Air losses in many places and one fire. There were steering problems in the many tight curves. Methane gas a minor problem, but an encounter with toxic waste in both headings caused a 9-day shutdown. Shields were slowed by an almost continuous deposit of boulders in one 386-ft long section. Progress was slowed further by unexpected encounters with wood cribbing obstructions and timber piles of abandoned piers. Operations and Maintenancea No problems were identified. Reaolution of Aasertions Re Subsurface Changus The contractor filed four claims for a total of $1,503,000. The claims covered extras for the toxic waste problem, the 386 ft of large boulders, the wood cribbing obstruction, and the timber pile obstruction. The owner accepted the contention that the conditions could not have been anticipated and reim- bursed the contractor by negotiating change orders amounting to $935,999. Analysis/Opinions Many miles of soft-ground interceptor sewer tunnels have been constructed by the owner, with substantially the same bidding format as the Red Hook tunnel. For this project the owner provided •boring logs• and a •geological report• which could be inspected or pur- chased by prospective bidders. 80 

The geologic report described the various soil strata identified in the boring logs, such as •fine sand-compact, till, possible boulders, etc. ,• with very little analysis or discussion of the effect of the varying geology on engineering and construction procedures and problems. It was not a •Geotechnical Engineering Report.• The soils were, in general, reworked glacial soils and the project ran parallel and in close proximity to a terminal moraine. There is considerable information available on the local geology and history as well as on numerous construction projects in the area, including 10 sub- way tunnels which crossed the line of the sewer tunnel. None of this was discussed in the report and few conclusions were drawn or evalua- tions JUde. The soil samples available for inspection were about 10 years old and of little help to the bidders. The major probleiiS of a geologic nature that affected construction of the tunnel were as followss • An excessive number of large boulders, many more than indi- cated by the boring logs, and sometimes occurring as large pockets with little or no fines. In one 400-ft length of tunnel, 166 large boulders were encountered and mined through. The largest boulder extended 13 ft along the axis of the tunnel. • Rock filled timber cribs (some noted on the geology report). • Pile foundations (not indicated on the borings). • Toxic chemicals and gases (not indicated on the borings or the geology report). • An area of very low cover under a heavily traveled industrial street, with major utilities and running sand. The construction problems encountered were severe, and delays were very costly as the tunnel was built in compressed air with six four-hour shifts per day. Fortunately, there was excellent cooperation with the owner and his contract manager, all with the attitude of how best to solve the problems and get the job done to the owner's specifications and requirements. Despite many substantial ·disagreements in negotiating claims for changed conditions, all disputes were settled during the course of the work through negotiated change orders and no claims were filed for liti- gation by the contractor. The major change orders relating to geolog- ical conditions totaled about $936,000 whereas the contractor had re- quested $1,503,000. (There was another major change order of $574,000 relating to special requirements of the Transit Authority while mining adjacent to more than five pairs of subway tunnels, but this is not re- lated to the purpose of this analysis.) A more complete •Geotechnical Engineering Report• would have pro- vided the contractors with more information for bidding purposes as well as for evaluating construction procedures. It might have predicted the incidence of boulders much more accurately (as a private report did). However, better data may or may not have resulted in a greater overall project cost to the owner; the size of the change orders were very nominal for the gravity of the problems, and in a competitive bid- ding situation the original bids might not have differed greatly. Con- 81 

tractors looking for work are notoriously (but not always wisely) opti- mistic about solving •field• problems. Sometimes they succeed and occasionally they do not. A difficult project like this could have be- come a catastrophe, greatly increasing the cost both to the contractor and the owner. It is neither fair to the contractor nor prudent for the owner not to provide all relevant information that can be obtained with- out excessive costs, including the geotechnical evaluation of the data as they impinge on design and construction of the project. 82 

CASB STUDY NO. 9 Naae of Project a Edward Hyatt Powerhouse (formerly Oroville Power Plant) Purpoaea Underground chamber for hydroelectric power production Locationa California (on the Feather River, 5 miles northwest of Oroville) Construction Perioda March 1964--June 1966 Site Investigation Perioda December 1959--october 1962 Sizea 550ft long, 139ft high by 71 ft wide (average). Project Costa Estimated $20,592,461 Bid $18,366,780 As Completed $42,414,628 Mined Tunnel Construction eo.ta Estimated $7,166,097 Bid $5,990,163 General Contract Mods $998,977 Subsurface Related overruns $16,300,000 As Completed $23,289,140 Subsurface Investigation Costa Not available s.-ary of Site Geologya Generally fresh, hard and massive amphibo- lite with some granitic gneissic zones. Three predominant joint sets with fractures, moderately to widely spaced. Many shear zones and schistose zones from 1 to 6 in. wide, containing crushed rock and clay gouge, dipping steeply and spaced 5 to 20 ft apart. Weathering along these zones, but not extending to powerhouse depth. Depth of overburden approximately 300 ft surface to crown. Water movement expected within fractures, joints, and weathered shear zones. Design Criteriaa Modulus of deformation of rock mass • 1.5 x 106 psi1 in situ rock stress determined to be isostatic at about 5,000 psi. Designed for relief of hydrostatic pressure (envelope grouting around the powerhouse with decreased injection pressures nearer the structure, combined with a system of gravity drains to relieve pressures on the structure) • Contract Provisionsa Type Unit price per cubic yd for excavation and concrete, per linear ft for rock bolts, and per pound for reinforcing steel. Estimated quantity variation limits not specified. 83 

Stipulationa Schedule and/or time of completion: 1,096 days for total contract. Definition of delay and suspension of work. Liquidated damages: $1,000 to $3,000 per day of delay and $100 per cubic yd for excavation outside the B line. Payment: Monthly, lOt retainage (optional after SOt completion). Construction method: Drill-and-blastJ full face in three separate headings in upper portion and quarry method in lower portion. Restrictions: None Disputes resolution: Decision by owner's engineer. If the contrac- tor disagrees, he may file a notice of potential claim; the formal claim must be submitted within 60 days. The engineer decides all claims and his decision is final. The only further recourse may be litigation. Geotechnical data aade part of contract docu.entaa Project geol- ogy report available on request, including summary boring logs and mappings in exploration tunnels, but no interpretation. Core sam- ples available for inspection on application. DisclaiMrsa OWner completely disclaims responsibility for, and accuracy of, subsurface data. Changed-conctitiona clauaea Yes Construction Metboda Drill-and-blast using truck-mounted drill jumbos (two platforms with six drills on truck bodies). Primary support of rock bolts and shotcrete with wire mesh. Conctitiona Bncountereda AB predicted by owner information. Probl... Bncountereda Conatructiona Extensive overbreak during excavation of benches near where adjacent tunnels enter the powerhouse. This required large quantities of rock bolts, steel ribs, and concrete backfill for stabilization. Rock movement in some areas and partial cave- ins of access tunnels. Operationa and Maintenancea No problems identified. Resolution of Assertions Re Subsurface o-angesa The contractor filed a $14,073,427 claim for the bench instability, contending that the com- plex design shapes were almost impossible to achieve in light of the ex- tensive network of shear and schistose zones. The owner denied the claim, maintaining that the joint patterns and frequency could be ob- served in the rock exposed in the powerhouse excavation and that the broken condition of the rock was due to poor blasting control and heavy blasting in adjacent tunnels. The owner forced the claim into litiga- tion. The Superior Court of California found in the contractor's favor within 6 months, but the $16,300,000 award (the amount claimed plus esca- lation and interest) was delayed by appeals until nine years after start of litigation. 84 

Analyais/Opiniona For a project in which the awarded amount from changed-condition claims was equal to 272 percent of the bid amount, the Edward Hyatt Powerhouse location was unusually well explored. Although the cost of the subsurface investigation is not available, the scope of the program appears impressive considering the extent of boring and seismic work, the amount of field and laboratory testing, the footage of exploratory drifts, and the peg model which was constructed. The inves- tigation indicated that construction would be generally within fresh, hard and massive amphibolite with relatively small amounts of granitic and gneissic rock. However, shear zones and some schistose rock were also identified~ it was predicted that between these two sources of in- competent materials there would be steeply dipping zones of crushed and/ or highly fractured rock every 15 to 20 ft along the chamber axis. Such zones did indeed occur, and the areas of poor rock caused se- vere shattering and over break in bench areas near intersections between the chamber and adjacent tunnels. The condition required unexpectedly large amounts of concrete backfill as well as additional rock bolts and steel sets for support of the excavation. The contractor contended that the fractured and sheared condition of the rock at the foot . of the powerhouse walls was inherently unstable and that the complex shapes re- quired in the large chamber were not possible to construct within the B Line. we must assume this contention to be factual because the courts eventually (after nine years) awarded the contractor the amount asked, plus considerable interest. The question then becomes: If the geologic site investigation was adequate to define ground conditions accurately, how did an almost un- constructible chamber configuration get into the contract documents? The answer would seem to be that the proper interpretation of geologic conditions as related to construction feasibility was not made by the geotechnical engineer or the designer. The effect of incompetent rock zones on the desired excavation outline apparently was not properly as- sessed during the design stage. A common tunnel design philosophy calls for the designer to size and space the elements of permanent support under the assumption that all temporary and initial support and the maintenance of a proper excavation outline are strictly within the pur- view of the contractor. This philosophy prevents the owner from improp- erly taking too much responsibility for routine field situations and operations. It may also obscure the need for geotechnical specialists and designers to maintain construction-wise staffs to review the plans from that particular point of view. such an approach can work well with small or uncomplicated open- ings, especially where the tunneling medium is well suited for under- ground construction. If that was the governing philosophy behind the Edward Hyatt design, it may have been inadequate because ground condi- tions were not ideal and the opening was neither small nor uncomplicated. Any powerhouse chamber is quite large, and the excavation shape and stress redistribution patterns are made complex by the intersecting tun- nels and the benches required for machinery emplacement. Planning for such a structure requires the designer to help ensure its ultimate in- tegrity by giving the greatest amount of thought to how the concept and the desired shape and dimensions can actually be executed in the field. This, in turn, requires that the designer and/or the geotechnical engi- 85 

neer (without usurping the contractor's final responsibility) review plans thoroughly for •constructibility• in light of the geologic situa- tion and make changes where necessary. Indications are that this step was not adequately pursued on the Edward Hyatt project, so there may have been a shortcc:aing in the final, interpretive stage, of the site investigation. 86 

CASB STUDY NO. 10 Name of Project• waste Isolation Pilot Plant Purpose• Exploratory access shaft to determine site suitability for storage of low-level nuclear waste. Location• New Mexico (approximately 30 miles east of Carlsbad) Construction Period• July 1981--December 1981 Site Investigation Period• 1974--1980 Sizea 2,242 ft deep1 11 ft 10 in. diameter. Project COsta Estimated $10,207,109 Bid $10,361,071 As Completed $10,113,904 Mined Shaft Conatruction COsta Estimated $6,977,207 Bid $7,419,705 General Contract Mods $ -171,388 Subsurface Related OVerruns $0 As Completed $7,248,317 Subsurface Investigation COsta Not available S.-ary of Site Geology• overburden consisting of 10 to 40 ft of windblown sand (approximately 20 ft at shaft location) underlain by silt- stone. Siltstone interbedded with sandstone and mudstone (the Dewey Lake Red Beds) overlies an anhydrite section interbedded with dolomite and mudstone which merges into the massive salt horizon (from a depth of 800 ft to greater than 2,400 ft). The salt horizon contains thin anhydrite interbeds and one zone enriched in potassium chloride. Design Criteria& Concrete key at 850-ft depth designed for lateral pressure of 75 percent of overburden weightJ steel liner and key de- signed for hydrostatic head of 600 ft. Contract Provisions• Typal Cost plus. (Drilling contract on •day work• basis.) Stipulations& Schedule and/or time of completion: Unknown Definition of delay and suspension of work. Payment: Monthly: 10• retainage until so• completion. Construction method: Blind hole drilling. Disputes resolution: Standard •General Conditions• for federal gov- ernment contract. 87 

Geotechnical data aade part of contract docu.entaa Vertical sec- tion (composite of two borings) included in contract drawings. Core samples available for inspection. DiaclaiMraa None with respect to owner-furnished information on subsurface conditions. Cbanged-conclitiona clauaea Yes Construction Method: Downhole drilling using drill derrick and hoist with 12-ft diameter rolling cutterhead. Permanent support of steel liner in upper 850 ft1 no final lining at greater depths, but with rock bolts and wire mesh for support as required. Conditione Encountered: As predicted by owner information, but less convergence than expected in salt. Probl... Bncountereda Conatructiona None of any significance. Operations and Maintenance: No problems identified. Resolution of Aaaertiona Re Subsurface Cbangeaa No assertions made. Analyaia/Opiniona The site investigation was can led out almost con- tinuously during 1974-1980 and covered an area of more than 100 sq miles before the final site was selected. The coat of this overall effort waa very high--in aggregate more than the cost of the shaft itself. I t is not possible to identify and separate those costs that are site specific to the shaft, but only a small percent of the investigation cost can be assigned to site description for design purposes. Two boreholes were drilled near the shaft site. Deliberately, drilling in the immediate area was held to a minimum so as to avoid possible c01m1unication path- ways into the repository area. Given that the project was conducted in a glare of publicity, much of it adverse to the concept of a low-level nuclear waste repository, it was essential that unforeseen problem& or delays did not occur. Any problem--particularly i f unexpected--would have been used as •proof• of site unsuitability. Thus, the preconstruction geotechnical investiga- tions and design were of necessity over-conservative. It was a classic example of •belt and suspenders• design. The skeletal design criterion was to rapidly construct an access shaft, plus a ventilation/escape shaft. The access shaft would be used to excavate chambers in the salt, at the preselected repository horizon, in which to conduct various long-term tests. There were two major specific design criteria. One was that the Dewey Lake Red Beds could not be allowed to become water saturated; his- torically, if wet the beds would swell, spall, and cave. The other cri- terion was that the minor-flow fresh water aquifer could not be allowed to contact the salt1 it would cut channels and could disrupt the shaft fittings in the unlined portion of the shaft (i.e., that portion in salt). 88 

Large diameter drilling was selected as the construction approach for several reasons: • It was demonstrably much faster, and there was no risk to personnel from working in a shaft bottom (no one entered the shaft until it was completed). • It was not subject to the delays and problems with water that have accompanied conventional shaft sinking in the area. • It provided minimum disturbance to the salt, e.g., no blast fractures, so that the necessary measurements of salt creep and long- term stability could be carried out in the shaft as well as in the chamber areas. • The unlined ventilation/escape shaft could be quickly and economically •slashed• (enlarged) to a size suitable for long-term usage, should the test program demonstrate acceptability of the site for a re- pository. In the meantime, the small shaft, which was a safety-dictated necessity during the test program, could be constructed very rapidly and at much less cost than a conventional drill-and-blast shaft. The construction manager developed the contract specifications and, because the technique and methodology had been preestablished by the owner, opted for a "day-work• type subcontract for the drilling opera- tions. (Equipment and personnel operated on a fixed hourly rate, with the rate dependent on the type of work being perfor1ned.) The construc- tion manager estimated the number of hours required for each category of work, and the drilling contractors bid hourly rates for their rig, an- cillary . equipment, and personnel based on their estimated quantities. The minimum size and capacities of the drill rig were specified in de- tail in the call for bids, and the drilling subcontractor's experience in similar work was also a bid appraisal consideration. The given geologic data consisted of a geologic column in the form of a strip log, with pertinent geologic and hydrologic comments in the margin. It should be noted that the local geology and hydrology were well known to the construction manager; therefore, with the type of con- tract, full details including geotechnical data were not essential to the drilling subcontractor. The construction method was blind shaft rotary drilling with cut- tings removal accomplished by a dual string circulation system. With this technique, a mix of high-pressure air and drilling fluid is pumped down the annulus between two coaxial strings of pipe (in this case 7 in. by 13 3/8 in.). The mixture flows into a chamber in the bit body, through jet nozzles in the bottom of the bit, and returns to the surface via the 7 in. inner pipe, carrying with it the cuttings from the hole bottom. A •blanket" of fluid, 150 to 200 ft deep, in the shaft prevents the air-fluid mix from filling the shaft. The conditions encountered were precisely as anticipated; the for- mation changes were within inches of where shown on the strip log. Aside from minor operational problems with the dual-string system, the construction proceeded as planned and scheduled. This project is not a good example of severe construction problems, or of highly critical geologic-geotechnical features. However, it is a good example of how smoothly construction can proceed when the hazard 89 

areas are recognized in advance and appropriate plans made for over- coming them. On this project, the major hazard by far was the tendency of the Dewey Lake Red Beds to absorb water, swell, and slough. Much of the drilling in the area for oil wells has been plagued by this problem, and many holes have been delayed or lost. What is an irritation in an oil well can be a catastrophe in a large drilled shaft. If the shaft walls collapse atop a •big hole• drilling assembly, the cost of the tools lost exceeds $500,000. In addition there is a delay of several months while new tools are procured. At the WIPP site, potassium ion was added to the drilling fluid to inhibit wetting of the shales, and the dual string technique minimized the exposure time. The project was completed ahead of schedule and under budget--a tribute to good geotechnical data, good engineering, and good estimating. 90