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.
8. Conclusions and Recommendations The basic objective of this study is to discover improvements in practice and procedures that will enable planning and conducting more effective geotechnical site investigation programs. This chapter presents the subcommittee's judgments on matters that bear on achieving the study objective. 'ltle conclusions drawn generally offer a view of current industry practice and areas that could be improved. In a few cases the conclu- sions are stmply observations of fact and required no particular analy- sis or deliberation. The reader will find same suggestions for changes in current practice--suggestions that are implicit in the way the con- clusion is stated. The recommendations offer the more specific statements on how the tunneling industry can generally upgrade subsurface investigations and expand their uses. Eight of the recommendations are firm proposals for policies that can be implemented within a short time. The others con- cern areas where research and development would benefit predictions and exploration techniques. The judgments presented herein are not all strictly verified by the data contained in the case histories. Some of the judgments were influ- enced by subcommittee experience and knowledge of projects that could not be documented in detail, but include many more than 87 projects over a 20-year time span. Even though the bases for development differed, all of the conclusions and recommendations are equally valid in the view of the subcommittee. In addition, although the subcommittee's study was confined primarily to mined tunnels at relatively shallow depth, the findings can be applied to most underground construction projects because the principles of subsurface investigation and contracting are so similar. 113
COHCLUSIOMS It is in the owner's best interests to conduct an effective and thor- ough site investigation and then to aake a co.plete discloaure of it to bidders. An owner has no legal duty to conduct a site investigation. However, if one is conducted, a variety of legal precedents would make the owner responsible i f actual site conditions are found to differ materially from those indicated by preconstruction subsurface explorations. Pre- contract uncertainty as to risk promotes increased costs for contingen- cies; post-contract uncertainty as to risk allocation fosters disputes and litigation. Disclaimers in contract docu.ents are generally ineffective aa a aatter of law, aa well· as being inequitable and inexcusable in 110et circua- stances. unexpected subsurface conditions are the primary cause of disputes and litigation arising from contracts for underground construction. The geotechnical investigation is a central element in the definition and allocation of risk. Disclaimers of information supplied are an inade- quate means of managing risk. It is the policy of most federal agencies and many owners to bear the risk of subsurface uncertainties and provide for differing site condition and changed condition clauses. Contracting docu.ents and procedures can provide for resolution of un- certain or unknowable geological processes or conditions before and dur- ing construction, rather than afterwards. The provision of clauses covering differing or changed conditions does not necessarily also provide a mechanism for prompt resolution of the issue. Adopting a baseline of risks (or a basis of geotechnical data) before construction would permit timely recognition of a contract change and provision for cost adjustment during construction, i f the conditions encountered vary materially. This should assist in reducing or eliminating contingencies for possible delays and disputes, and lead to more realistic cost estimates and more competitive bidding. On aajor projects especially, it is important that (a) the owner .-plo, a multi-disciplined teu including engineering geologists, engineers, and a construction specialist to develop subsurface data and evaluate their t.pact on design and construction, (b) designers and geologists ~seas a thorough working knowledge of construction â¢thode and equi588nt so that the proper geotechnical data are secured and design is consistent with construction syst... J and (c) contractors emplo, geologists experi- enced in underground work to evaluate and interpret the data provided at the ttme of bidding, thus ensuring that all the inforaation obtained is fully considered in preparing bids. The most extensive and effective geologic site investigations are of limited value if not incorporated fully into the design, estimating, and bidding processes. Too often, the significance of geologic site condi- tions is not emphasized appropriately in siting, budgeting, and design. Important information either may not have been considered due to poor 114
communication between various disciplines, or may never have been ob- tained due to a failure to recognize the need. Current and developing underground construction methods demand greater attention to the collec- tion and application of geologic information. The exploration programs, interpretation techniques, and even the potential of the investigator(&) to conduct an effective investigation must be evaluated. It is essential that the owner, designer, and contractor know when additional skills or knowledge may be required. It is the user's responsibility to help ensure that geotechnical investigations reduce, rather than contribute to, risk and the incidence of unanticipated adverse conditions. Site investigations have to proceed through, but should not always end with, co.pletion of the feasibility/alignment setting/final design pro- graJIS. Owners must recognize that the preconstruction site investigation should be an iterative process. A project comprises several phases, and appropriate data must be collected and analyzed to support the require- ments of all phases. Anomalies should not be left unresolved by the presumed â¢final⢠program, but further explored by another program, and then another, if necessary. All geotechnical data that an owner can sustain economically should be developed. This philosophy should extend to developing additional information when it is important for good bids, even if the information no longer is directly relevant to the design itself. For example, an easily performed but generally ignored investi- gation procedure is the continued reading of groundwater levels in ob- servation wells as long as there is time to print the information for use by bidders. There is always the possibility of a late-developing change in the groundwater table having major ramifications for construc- tion operations. Moreover, it is not too late to continue exploration after a project is already let for bid7 bidders may require data that entails additional exploration. In some instances, post-bid and even post-award investigations may be justified. Procedures for logging, docu.enting, and preserving saaples froa bore- holes require improvement. BOreholes should be observed and logged by experienced engineering geologists. Modern drilling techniques and equipment should be used to allow optimal core recovery. Color photos of all cores should be taken soon after removal from the borehole in order to document the condition of the cores at the time of drilling. Cores frequently deteriorate with time1 samples are removed for testing and, through handling, are mixed up or disturbed. Efforts should also be made to preserve cores until at least the completion of construction. Permanent retention of the cores at the project site or an associated facility would be the most desir- able approach. For cores that deteriorate rapidly, special preservation techniques such as wrapping in plastic or sealed tubes may be necessary for adequate preservation. Soil sampling procedures are also in need of improvement and stan- dardization. For example, soil sampling should be essentially continuous through the level of the planned tunnel. Use of high torque equipment such as rotary drilling or hollow-stem augers should be restricted in overburden, particularly below the planned crown of the tunnel. 115
Geophysical methods can be used to advantage, especially in coordination with boreholes. Geophysical methods have the potential to greatly expand knowledge of the subsurface when used to interpolate between boreholes. Many geo- physical techniques are not widely used or applied to construction proj- ects, but are worthy of continued investigation and development. Seismic refraction surveys profiling the rock surface between boreholes help eliminate the problems associated with high rock in the invert of a soft- ground tunnel and soil intrusion in the crown of a rock tunnel. Other techniques such as resistivity, gravity and magnetic survey can be used to identify anomalies where borings should be made. Ideally, geophysi- cal surveys should be performed prior to drilling the final design bor- ings to allow optimum placement of borings to check different conditions indicated by the geophysical surveys. Groundwater and its effects on the subsurface aaterials Mrit greater attention in exploration programs. The presence of water accounts, either directly or indirectly, for the majority of construction problems. Most major tunnel projects should have one or more long-term pump tests, executed in accordance with good standard practice and conducted so as to test the various formations and conditions to be encountered during construction. These tests should include observation wells to directly observe pumping effects, as well as drawdown and recovery. Chemical tests of groundwater should be per- formed on a routine basis. Recent advances in computer modeling of groundwater flow may have applications in improving the ability to pre- dict flow into the excavation, and thus are worthy of investigation. Laboratory testing of the subsurface aaterials generally needs to be increased. Experience has shown, for example; that in rock tunnels at least 50 to 60 unconfined compression tests for each significant lithologic unit are necessary to adequately characterize the range and means of strength values. Silica content is rarely determined in testing programs, yet it is an important parameter in allowing the contractor to predict advance rates and abrasive wear on equipment. In the same vein, sufficient and careful testing of overburden and soft-ground material is important. Truly adequate testing calls for supplementing standard split spoon sam- ples with undisturbed samples from each stratum or zone that affects the tunnel. It must be noted that testing of disturbed rock or soil samples places severe limitations on the value of the resulting data. Exploratory adits and shafts are generally justified only when abso- lutely essential to obtain critical design data or when a substantial benefit to construction is indicated. These exploratory techniques are very expensive and are of question- able cost-benefit in many cases. In some cases, pilot tunnels have actually increased problems during construction of the project; misalign- ment or exceptionally poor work in the adits or shafts may increase the cost of the final opening. An alternative view is that a significant portion of the pilot tunnel or shaft may be charged to subsequent work if the final opening incorporates the pilot tunnel or shaft. Generally, 116
however, the money expended on an exploratory adit or shaft may be used more effectively for additional boreholes, groundwater investigations, laboratory testing, or engineering evaluations. Maintenance of technical data obtained during design and construction of underground projects often is not pursued by owners or demanded of their consultants and contractors. A surprising quantity of exploration, design, and construction data is poorly recorded, filed without easy access, lost, or discarded by owners, construction contractors, and others. In conducting this study, the subcommittee found that records for older projects, as well as for some more recent projects, were often difficult to locate or impossible to obtain. This was because they had either been stored in a manner that discouraged file searching, or simply destroyed. For newer projects, the difficulty in locating information was generally caused by poor rec- ordkeeping. Although this was more often the case for agencies involved in only one construction program, records were not always reasonably available for reference from agencies that build and operate tunnel after tunnel. Experience has shown that relatively few major underground proj- ects fail to develop problems during their operational lifetimes. In many cases, data obtained in the exploration, design, and construction phases of the project are essential to defining the cause of the problem and the best method of correction. If records are not available, the data must be obtained again and the procedure can be time consuming and costly. The difficulty and expense involved in securing suitable data can sometimes lead to inadequate or even â¢patchwork⢠solutions. RBCOIMEHDATIOHS Expenditures for geotechnical site exploration should be increased to an average of 3.0 percent of estimated project cost, for better overall results. 'l'he low level of expenditure typical of current practice does not correlate well with estimated and actual costs or with construction problems and claims. overall, increasing exploration can be expected to decrease the incidence and severity of construction difficulties and eliminate a significant portion of the extra costs associated with unan- ticipated geologic conditions, including project delays, claims, and litigation. Increased explo.,:ation should lead to more reliable engi- neers⢠estimates and owners⢠budgets, as well as more accurate bids. It is possible that increased exploration would result in higher engineer's estimates and higher owners⢠budgets, thereby reducing the direct cash savings resulting from fewer claims. However, savings still would ac- crue from eliminating attendant delays, lawyers⢠fees, and hidden costs. The level of exploratory borings should be increased to an average of 1.5 linear ft of borehole per route ft of tunnel alignment, for better overall results. Current boring practice is not consistent with the evidence that boreholes are the best single exploration technique for providing 117
reliable information to designers and contractors. Borings provide actual physical samples for direct observation and testing, a feature that makes them less subject to misinterpretation than more indirect methods. However, some factors (including the great depths and/or dif- ficult surface access of some sites) prevent this investigation tech- nique from being given the intensive use it merits. Exploration at 0.6 lin ft of borehole per route ft generally initiates a decrease in the deviations between the final tunnel cost and both the bid price and en- gineer's estimate. However, an increase to 1.5 produces more desirable results. Beyond this level of effort, the risks of geologic uncertain- ties, although not eliminated, may be reduced to the point of diminish- ing returns for borehole footage drilled as a matter of general practice. The optimum level for borehole footage entails an increase higher in magnitude than the optimum level for exploration expenditures, but the recommendations are not incompatible. A substantial portion of the cost of any drilling program is devoted to initial mobilization, and the more modest programs incur maximum charges per ft of borehole. However, as the number and/or depth of boreholes increases, the unit prices flatten out and even decrease. Moreover, the cost of the overall exploration program includes expenses for report writing and other miscellaneous items which do not rise in direct proportion to borehole footage. The owner should make all his geotechnical information available to bidders, while at the sue tbae eliainating diaclabaera regarding tbe accuracy of the data or the interpretations. In the past there has been a tendency among owners to give bidders as little of their interpretive information as possible in order not to be held responsible for any mistakes made in extrapolation from hard data. Owners would make available the logs of boreholes--because they are presumably factual--but withhold the geologic reports because of their interpretive nature. The result was that various contractors were bidding on different bases, depending on their personal experience or access to knowledge apart from the boring logs. Bidding contingencies tended to be high to cover the construction unknowns. This situation is undesirable and can be mitigated significantly if the owner will present all the geotechnical information, and without disclaimers. The owner bears some responsibility for errors in the subsurface predictions, but it creates a fairer bidding atmosphere and can ultimately lower construc- tion costs. All geologic reports should be incorporateCS as part of the contract documents. Some owners follow the philosophy of making all of their subsurface data available to bidders, but not making it a binding contract document. The material is presented for examination, yet not provided or sold with the contract drawings and specifications. Geotechnical documents made available in this manner are often accompanied by a disclaimer stating the owner will not be held responsible for any interpretations or use made thereof. One consequence of this procedure is that some bidders may not rely on the information in spite of ita possible accuracy and may not plan their construction operations with all salient facts in mind. A second consequence is that if litigation over changed condition 118
claims is instituted, much time can be spent in arguments over whether the geologic information (or misinformation) can or cannot be blamed on the owner. The procedures should be simplified, even though the owner will then be more surely liable for any errors in interpretations. The result--more consistent and accurate bids--will be worth the added responsibility. Designers of mined tunnels should compile a â¢Geotechnical Design Re- port, ⢠which should be bound into the specifications and be available for use by bidders, the eventual contractor, and the resident engineer. A geologic site investigation is generally completed by the middle stages of design and, therefore, the geologic report cannot comment on many of the late-developing plans worked out by the designer. As a result, bidders are uninformed on many important design/construction matters that may have been given serious consideration prior to the let- ting of bids. The goal of the Geotechnical Design Report should be to explain the geotechnical rationale for design and the anticipated effect of geology on construction. Such reports should result in much better informed bidders, improved construction procedures, and probably lowered costs associated with a reduction in bidding contingencies and changed condition claims. The WMATA Geotechnical Design Reports (Appendix D) illustrate standard items that should be described in such reports. In addition, including one of the systems for rock classifications (e.g., RSR, RMR, Q-System, or Terzaghi) may be useful, provided that the system is applied properly. Monitoring of aabient conditions prior to construction should be under- taken to establish a baseline of information for comparison during and after construction. Records of specific data can be useful in preventing or settling disputes related to construction conditions or effects, as well as in protecting both owner and contractor from frivolous claims. The process can range from visual inspection of structures within a zone of vibra- tion, to a detailed survey of existing damage in adjacent structures, to long-term measurement of groundwater levels. For construction in rock where drill-and-blast procedures are applicable and sensitive structures exist at the site or nearby, preconstruction blast/vibration/noise/sen- sitivity measurements should be made to compare with later effects and to use in establishing a public relations program. A crack survey, ele- vation benchmarks, and vibration measurements of non-construction activ- ities should also be undertaken. Pre-bid conferences and site tours should be conducted to ensure that all bidders have access to the maximum amount of project information. The end result of a subsurface investigation should be to place as much geotechnical information as possible in the hands of bidders. A good site tour can help accomplish this by allowing bidders to get the â¢lay of the land⢠and see the physical features of the project for them- selves. However, such tours may lose some of their effectiveness if not conducted by a knowledgeable owner representative. Those bidders not familiar with the territory or the project can miss important features by being left to discover everything for themselves, and bidder ignorance 119
is not in anyone's best interest. In cases where a test adit or shaft has been constructed, the conducted site tour becomes a matter of even more concern. In the same way, a pre-bid conference is a good way of assuring that all potential bidders have an opportunity to clarify any confusing issues in the contract documents. Some owners choose not to spend time with such conferences because attendees tend to raise few issues for fear of revealing to competitors their amount of knowledge or their bidding strategies. This situation should not deter owners from making the effort. A conference should always include an oral geotechnical brief- ing by the project designer. This feature is especially important where some policy or circumstance has made it difficult for bidders to obtain the appropriate geotechnical reports or boring information. In addition, the bidders⢠responses to that briefing can assist the owner in evaluat- ing the effectiveness of the site investigation. Geologic information from preconstruction explorations and as-built tunnel mapping and construction procedures should be compiled in a report detailing project completion. It is rare to find wrap-up reports that describe the mapped tunnel geology and construction procedures, even among owners who build tunnel after tunnel. Without such a report, there is no formal way for an owner to confirm geologic predictions and find out where assumptions were right or wrong. There is also no easy way to resurrect records of operations and apply the experience to future projects in order to avoid the repe- tition of errors. Such information can be invaluable in the event of damage to or malfunction of the tunnel during its operational life. There are cases where post-construction problema (e.g., drain clogging, lining distress) were difficult to diagnose and correct because actual construction (or geologic) conditions were not recorded. As a minimum, the as-built geotechnical conditions should be reviewed by the original exploration team. If practicable, the original team should assist in the post-construction mapping. It is in the owner â¢a interest to create such a record for improving design, contracting, and construction manage- ment techniques. Expense would be involved because the â¢as-built⢠report could approach the size of the original design report, but it would be to the owner's long-term economic benefit to engage in the effort. Investigation methods and predictions should be improved for three spe- cific conditional in-situ stress, stand-up tt.e, and groundwater. In-situ stress is one of the conditions not always adequately pre- dicted by designers. A .better understanding of the geologic history of the site is needed, e.g., eroded cover, normal variation of rock strength, tectonic activity. However, merely paying more attention to the situation during exploration might not be as effective as hoped, because the instruments and predictive techniques need further develop- ment. Research is especially needed for predicting stresses at great depth (more than 1,000 ft), particularly when coupled with below average rock strength (less than 6,000 psi compressive strength). Estimates of stand-up time developed from information available prior to construction are usually indefinite (or not provided). Reliable esti- mates are important for design and bidding. Stand-up time is a major 120
consideration in selecting appropriate construction methods, equipment, and support system. The Rock Mass Rating (RMR) system may show promise here. In addition, RQDs (when properly determined and recorded), coupled with close inspection of joint and fracture conditions (roughness, fill- ing materials, degree of continuity, spacing, and amount of opening) are useful tools of a semi-quantitative nature. The occurrence, behavior, and effects of groundwater account either directly or indirectly for the majority of problems encountered in under- ground construction. This situation is a strong indicator of the need for research and development. First, there is a lack of good quality field pump tests--pump down with observation wells, along with recovery tests. Second, there is inadequate understanding of the effects on ground stability that can result from even a small amount of water flow. In rock tunnels, small quantities of water can substantially reduce fric- tion along joint surfaces; its exit pressure can dramatically affect otherwise stable rock. Water can also cause swelling and induce squeez- ing in certain types of rock. Development of a data base would assist in sorting and evaluating the complexities of the problems presented by groundwater. One effective and relatively inexpensive way to establish a good data base would be for owners and contractors to begin keeping careful records on quantities of groundwater flowing into the various reaches of tunnels during excavation. Currently, such data are recorded on an irregular basis, and thus much valuable information is irretriev- ably lost. Ideally, the records should be supplemented with notations regarding the nature and extent of any problems and the effects on con- struction. Illproved horizontal drilliDCJ techniques should be developed that can recover rock core and penetrate lODCJ distances without wanderiDCJ from line and grade. The need is especially severe for tunnels beneath mountains where, except for portal areas, difficult access and/or great depth generally make the necessary number of vertical boreholes prohibitively expensive. The ability to core drill accurately from a portal and along the tunnel alignment would help investigators to determine not only what is there, but also the true boundaries and thicknesses of geologic features as they would ultimately be encountered in the advancing excavation. Research and development should be conducted to expand the capabilities of geophysical or other remote sensing methods for obtaining geotechnical data between boreholes and from the surface down to depths too great or too costly for boreholes. Although boreholes provide the best kind of geotechnical information from within their own confines, interpretation or extrapolation is essen- tial to project that knowledge to some useful distance beyond the bore- hole. A higher degree of interpretation/extrapolation is required to glean information from depths too great for economical borehole penetra- tion. The process can be greatly abetted by reliable techniques of geo- physics and remote sensing. However, in comparison with some industries (e.g., petroleum exploration), engineering investigations make minimal use of these more indirect methods of data collection. A major reason is their relative lack of preciseness, which can lead an owner to the 121
uncomfortable perception that the data are readily subject to more than one interpretation. Considering the ability of remote sensing techniques to cover continuous extents of ground, subsurface investigation would be vastly enhanced if those techniques could be developed to the point that their results were as reliable as borings and trusted equally by both designers and contractors. It should be noted that some federally financed research on deep remote sensing methods is being conducted; the methods are showing promise but still require extensive testing to prove dependability. 122
