North American Tunneling Conference, April 17-21, 2004
The Sequential Excavation Method (SEM), also referred to as the New Austrian Tunneling Method (NATM) is becoming increasingly popular in the US for the construction of shafts, tunnels and other underground structures. Besides the economic competitiveness of SEM, advantages of the method include the outstanding flexibility in terms of geometric shapes and its adaptability to varying ground conditions, as well as the substantial reduction of impacts and disruptions to communities in urban areas. Even though used in the US since 20 years, it is still considered a new method grasping for acceptance and support in the engineering as well as the contracting community. To take best advantage of SEM and to best utilize its cost-saving poten-tial strategies have to be developed that include establishing of contractual frameworks, procurement strate-gies and unit price contracts, employment of skilled and experienced designers willing to take responsibility for a prescriptive design, extending the designer’s responsibility and involvement into the construction period and allowing for a competitive bidding environment. Technical pre-qualification of contractors needs to be reviewed on a case by case basis and should only be required if varying ground conditions dictate frequent change of support means, application of more sophisticated additional support measures like SEM Toolbox items and utilization of complex excavation sequences with multiple headings. Some of the strategies extend beyond what is common practice, and therefore require flexibility and adaptability of owners as well as contractors to deliver a successful project. The Beacon Hill Light Rail Station in Seattle is a recent example where all these issues have been considered and implemented.
About 20 years ago, in the late 1970s and early 1980s the first applications of the New Austrian Tunneling Method (NATM) for tunnels in Pitts-burgh and Washington, D.C. generated attention within the US consulting and construction industries (Martin, 1984). The new method to construct tunnels resulted in low bids either in a competitive bidding environment (bidding on two alternative construc-tion methods) or as value engineering by contractors (Cavan et. al. 1985). Due to the success of the NATM value engineering solution the Washington Metropolitan Area Transit Authority (WMATA) de-cided to use the method either as alternative or sole design on their upcoming contracts (Heflin, 1985).
Since then a variety of projects have been de-signed and constructed using NATM, a method re-cently also referred to as Sequential Excavation Method (SEM), including transit and highway pro-jects in Dallas, TX, Boston, MA, Jersey City, NJ, New York and Seattle, WA. Modifications to the original concept were required to adjust to the do-mestic design and construction environment. These include the more robust base design, peer reviews, pre-qualification of contractors, strict requirements for construction management and supervision, the careful evaluation of proposed means and methods, application of risk management and the adjustment of the contracting strategy.
For the method to be successful and to best utilize its flexibility and cost savings potential a series of requirements has to be fulfilled. Owners have to adapt and accept the concept of the method that is labeled as “innovative” and “new” by demonstrating willingness to reach beyond well established boundaries in terms of procedural and contractual frameworks. Skilled and experienced designers shall be employed to continue the optimization process and provide economic design solutions combined with a robust contractual framework. Finally, due to the lack of experience of most American contractors especially with soft ground mining methods in urban areas and the importance of observing the ground and its movements, the construction has to be supersupervised in a way to assure that the design intent is conveyed into the construction phase.
NATM was developed in the 1950s with shot-crete being an integral part to stabilize openings immediately after excavation of each round. Initially developed for rock tunnels, the method was ad-vanced in theory and practice and adapted for soft ground tunnels in urban environments, with the first application in Frankfurt, Germany in 1968. Since then NATM was used throughout the world on many projects for transportation, water/wastewater conveyance or other purposes.
Due to the modifications of the method to adjust it to the US market the term Sequential Excavation Method (SEM) is sometimes used instead of NATM. Such modifications include the development of “robust” design as a base case supplemented with tool-box items, instead of following the original idea of an observational method, where support measures are selected based on geology at the face and deformations of the ground and lining. However, the general principle of mobilizing the surrounding ground by developing the maximum self-supporting capacity of the rock or soil remains unchanged.
“Greater skill is needed to avoid (minimize) ground load than to resist it” (Rziha, 1872), a statement from a tunneling engineer of the 19th century is still relevant and valid today and separates experienced designers from inexperienced incumbents in the field of SEM design. It is not an achievement to turn soft ground into concrete before mining a tunnel and thereby incurring substantial costs. The challenge and satisfaction of a SEM designer shall always be the development of the most economic, but of course safe design for all tunnel structures, considering all relevant issues, including geology, overburden, settlement sensitivity and technical feasibility.
Reference is made to a series of publications where in-depth information on the principles of NATM/SEM for rock and soft ground are described in detail.
One of the early decisions an owner has to make is to decide on the procurement strategy. Assuming that no participation in funding of the project by the contractor is required (i.e. Public Private Partner-ship), there are generally three contracting options available:
The aim of the procurement strategy is to achieve the optimum balance of risk and control, depending on the legal environment, where the project is lo-cated. The procurement route should ensure that de-sign, construction, operation and maintenance are considered as a whole, and that the delivery team for all of these aspects works together (OGC, 2003).
The traditional design-bid-build contract, where a single contractor acts as the sole point of responsi-bility for the management and delivery of a con-struction project on time, within budget and fit for the purpose for which it was intended is the most widely used and suitable form of procurement for underground construction, especially where ground conditions are difficult, and the owner has high ex-pectations on the quality of the final product. Com-pensation for the contractor is accomplished with lump sum or unit price payments or a combination of both, depending on the complexity and predict-ability of the project.
SEM project experience has shown that some ad-justments are required. Unit price contracts are gen-erally preferable to maintain the flexibility and adaptability of the design to the actual ground condi-tions encountered. It was also found valuable to ex-tend the services of the SEM designer into the con-struction period due to the detailed knowledge he has about ground, geotechnical and geo-hydrological conditions and the intensive design development re-quired for a thought-through SEM design. The design philosophy and design intent are thereby con-veyed into the construction phase.
Design-build can be defined as “using a single con-tractor to act as a sole point of responsibility to a public sector client for the design, management and delivery of a construction project on time, within budget and in accordance with a pre-defined output specification using reasonable skill and care” (OGC, 2003). There have been many attempts throughout various countries to make the design-build concept work for the construction of under-ground structures. While the approach seems rea-sonable and has been applied for many projects with success, underground construction has one major specific characteristic, namely the ownership of the nature and behavior of the ground. Differing site conditions can result in claims of large magnitudes with lump sum contracts due to the lack of detailed information about the contractor’s bid.
Risks of the design-build contract include the transfer of too much or inappropriate risk, which might not be cost effective. Also, the clarity of the output specifications that might be based on assump-tions that are inaccurate represent the potential for major cost increase in the case of changes or ad-justments of these specifications.|
A strong and powerful construction management with a knowledgeable SEM support team on the owner’s side, and an experienced contractor are es-sential for successfully completing mined tunnel projects on a design-build basis.
This procurement option also known as time and material (T&M) contract is rarely used for the con-struction of mined tunnels due to the high degree of responsibility and risk that remains with the owner. A highly experienced team on all sides, the owner, designer and contractor are required for successfully completing projects using this approach.
One successful application of this concept was the construction of seven cross-over caverns at the Exchange Place Station in Jersey City, NJ (Dinkels, 2003). The owner, the Port Authority of New York and New Jersey was under high pressure to rehabili-tate and reopen the station affected by the collapse of the World Trade Center in New York and re-open train service between New Jersey and New York for the millions of commuters at a specified date. The schedule driven project did not allow any of the tra-ditional contracting methods, therefore the cost plus fee contract was selected. Close coordination be-tween the owner, the design team and the contractor combined with a very experienced and powerful construction management resulted in the on-time completion of this project.
The owner requested from the contractor a fixed fee and a lump sum fee for its project management. A substantial incentive was provided to finish the project on or even before the specified completion date, and no penalty was foreseen for late comple-tion. This system generated a team-like environment where every party had the same goal and focus.
Designing SEM tunnel structures is different from designing other structures due to the complexity of and the interaction with the surrounding ground, when considering stress redistributions and devel-opment of equilibriums after each excavation and support step. Excavation sequences, pre-support, face support and ground improvement methods have to be developed depending on the size of the struc-ture and the ground conditions.
A comprehensive subsurface exploration program is essential and becomes more critical the more com-plex ground conditions are. An essential part of the investigation is the determination of the hydrologic regime, which can comprise of multiple groundwa-ter horizons. In some cases, especially when the ground has been affected by seismic or tectonic ac-tivities it will be impossible to determine the exact stratification of the various subsoil layers. The variability has to be taken into consideration during the design.
While it is typical to develop excavation and sup-port classes based on various available classification systems for hard rock, soft ground is usually more variable and results in mixed face conditions of some sort. Traditional classes are therefore difficult to define. A different concept is required to maintain flexibility but also to assure proper support during all phases of construction. Therefore the “SEM Toolbox” concept has been developed to cope with all, including the most adverse conditions.
Developed for NATM tunneling in soft ground, the SEM Toolbox concept follows the requirement for multiple or selected pre-support, ground improve-ment and support elements to stabilize the open face of an excavated tunnel heading. The complete tool-box contains all elements to allow mined tunneling through virtually any ground (Sauer, 2003). For each project, in addition to standard support measures toolbox items are selected that are required for the expected ground conditions and contingency proce-dures. The toolbox items are defined in the contract documents in terms of material, installa-tion/application procedure and installa-tion/application criteria.
Figure 3 shows a typical cross section and longi-tudinal section for an SEM tunnel during excavation. SEM Toolbox items are numbered and described be-low.
Standard support measures typically include fiber reinforced flashcrete (9), reinforced shotcrete (2) and standard dewatering measures.
SEM Toolbox items include the following:
Sidewall Drift Improvements:
Pre Support Measures (3):
Face Stabilization Measures:
Ground Improvement Measures:
While being in conflict with the more traditional lump sum payment method, the SEM Toolbox con-cept requires the use of unit prices for the various items selected as part of the project toolbox, while standard support measures are still compensated with lump sums. Quantities for the toolbox items are estimated and form the basis for the bid. Standby quantities are defined which must be available on site ready to be applied at any time during excava-tion of the respective SEM tunnel. Actual quantities will vary and depend on the actual ground condi-tions encountered. Application and installation of the toolbox items shall be approved or directed by the Engineer.
For the contractor it is important to have criteria for the installation/application of each of the toolbox items as well as the estimated quantity for each tun-nel reach. For this purpose typical tables are pre-pared that define relationships between Ground Type and SEM Toolbox Item (see Table 1), and Tunnel Structure and SEM Toolbox Item Quantities, which should be part of the Geotechnical Baseline Report.
Table 1: Relation between Ground Type and SEM Toolbox
Instrumentation and monitoring of ground move-ments and lining deformations is essential for suc-cessful SEM tunneling. Instrumentation comprises surface monitoring, ground instrumentation, and in-tunnel instrumentation, to monitor lining perform-ance. Surface monitoring includes monitoring of surface settlements and building deformations, and can require surface settlement points, tiltmeters, crack monitoring devices and vibration monitoring, depending on the sensitivity of the overlying and ad-jacent structures. Instrumentation for the surround-ing ground includes inclinometers, extensometers and deflectometers. The purpose of these instru-ments is to determine the loosening zone in the ground, and to observe movements of the tunnel face by monitoring the ground ahead of the excava-tion. In-tunnel monitoring includes convergence and lining deflection measurements, and pressure and stress monitoring by utilizing ground and shotcrete pressure cells.
For the ground and in-tunnel monitoring the read-ing frequency and reporting procedures are essential. Timely reporting to allow interpretation of data is the base for decisions at the face regarding the ade-quacy of the support, round length and sequence of installation. While monitoring of surface and ground deformation instruments is carried out from the sur-face and can also be done remotely, in-tunnel moni-toring disrupts the excavation and support process significantly. Today, optical surveying methods are used that are less disruptive to operations but still require an instrument to be placed in the center of the drift or tunnel.
The decision of who is carrying out the readings requires some consideration. Because it is a disrup-tive activity it is recommended to make it a respon-sibility of the contractor. This option allows the con-tractor to stage his work such that monitoring is a part of his excavation and support sequence. Two options are feasible under this approach, that the contractor either reads himself, with qualified survey personnel meeting or exceeding the specified re-quirements, or that he employs an independent, qualified surveyor. When tunneling underneath sen-sitive buildings an owner might be more comfort-able specifying an independent surveyor. The con-tractor is responsible to submit the raw and processed data to the owner in a pre-determined time frame, where raw data should be submitted immedi-ately after they become available, and processed data should be available no later than 24 hours after reading.
Alternatively, all surveying could be under the owner’s control by employing a surveyor to read the in-tunnel instruments. This approach requires close coordination between the surveyor and the contrac-tor and often results in access problems to carry out the reading. Damaged survey points, muck piles and parked equipment are some of the obstacles to over-come. Therefore, this approach is generally not rec-ommended.
Unique in the American way of design and construc-tion is the high degree of performance elements in a traditional final design package rather than provid-ing a prescriptive base design, where contractors can propose alternatives and submit value engineering proposals. The concept behind this philosophy is that final decisions on means and methods of con-struction remain with the contractor. This should guarantee most economic solutions for mainly tem-porary construction elements where contractors have more knowledge or provide the flexibility to choose the most preferable solution based on the contrac-tor’s experience.
In case of SEM tunneling, designers are usually more knowledgeable in terms of ground conditions and behavior at the time of contract award due to the extensive exploration and research during the design phase and the resulting development of excavation sequences based on experience and structural analy-sis. Leaving the design of the temporary support, in this case the ground improvement, pre-support and initial shotcrete lining to the contractor is not rea-sonable in the short time frame allowed between ad-vertising the project and start of construction. In ad-dition, contractors can usually not refer to the extensive SEM experience designers provide due to their involvement in projects worldwide.
To reduce risk of cost and schedule overruns the following elements of the SEM design should be prescribed:
The prescription of ground improvement and pre-support elements is only possible to the extent that the anticipated methods are specified and quantities are estimated, but final decision on location and ap-plication will depend on ground conditions encoun-tered and will be determined in the field.
Waterproofing is prescribed due to the many negative examples and leaking tunnels. So far, only the membrane waterproofing system has proven to provide watertight structures.
Despite the high degree of prescriptiveness there is still flexibility for the contractor to optimize his operations on site, such as elements like the con-struction sequence (sequence when various tunnels are built), the advancement of split headings (only minimum distance is specified), utilization of its equipment, selection of wet or dry shotcrete and - with some limitations - the selection of SEM Tool-box items by proposing elements of choice (Sauer & Gold, 1989).
Prequalification of contractors for tunneling projects has become popular in the last 10 to 15 years to re-duce risk. However, the process does not always provide the expected outcome. Problems encoun-tered include that bidders are pre-qualified based on previous successful projects but the crews have left the company in the meantime, the lack of similar past work experience when new methods are em-ployed, and the difficulty to provide resumes for key personnel sometimes a year or more ahead of the ac-tual construction.
There is no question that bidders should be quali-fied to perform the work, but how can this be ac-complished and what/who needs to be prequalified? Upon other criteria responsibility, financial viability and size of the prospective bidder as well as previ-ous experience with similar projects, including key personnel should be evaluated. Tunnel projects are usually singular projects and only few agencies or owners have historic data on successful or unsuc-cessful contractors and procedures. Also, criteria that would eliminate a specific contractor based on a problem on a previous project are difficult to formu-late and execute. It has to be recognized that Pre-qualification does not relief any owner of the re-sponsibility to monitor and supervise construction (Brierly, 2003).
In the US, there is still a lack of experienced con-tractors when it comes to large underground struc-tures to be constructed in soft ground. The two schools of thought in this case are to qualify all con-tractors based on financial capability and responsi-bility, and have a strong, SEM experienced con-struction management team available to guide the contractor through the process, or to rely on exper-tise from abroad. In the first case a team-like envi-ronment has to be established that allows the con-struction management to train, guide and work closely together with the contractor heading for a common goal. Alternatively, the contractor could add SEM experienced key staff that would fulfill the same role and train and guide the contractor through the project. One example of a successful application of this arrangement was the widening of the Berry Street Tunnel in Pittsburgh, PA. The contractor with no experience in tunneling, shotcrete or NATM/SEM decided to propose a value engineering alternative utilizing NATM for the tunnel excavation and support. He was supported by experienced NATM engineers and superintendents from his de-signer. The Port Authority of Allegheny County ac-cepted the proposal and the project was completed ahead of schedule and generated cost savings of $2 million (Garrett, 1998).
In the case of the Beacon Hill project in Seattle the owner, Sound Transit decided to increase the prequalification requirements requiring the US con-tractors to most likely look abroad for the requested expertise. The decision was made based on the fact that the tunnels to be constructed would be the larg-est SEM soft ground tunnels to date in the US and the station construction is on the critical path of the entire project, the Link Light Rail - South Link.
The above mentioned Beacon Hill light rail station is the most recent project, where large tunnels in soft ground are designed using SEM. During the design development many of the above concepts were re-fined and applied.
The complex geology, with fractured, inconsis-tent glacial deposits and multiple water horizons re-quired the utilization of a series of excavation se-quences, ground improvement and pre-support measures in order to provide a safe design. Details about the design of this underground station are pro-vided in another paper of this conference (Laub-bichler, et. al., 2004).
It is important to mention that during the design the SEM Toolbox approach was further developed and allowed an economic design approach for the conditions at hand. A strict prequalification was car-ried out to select qualified and experienced bidders able to utilize the designed tools for a safe excava-tion.
The approach for the instrumentation and moni-toring is different for surface and in-tunnel instru-mentation. For the in-tunnel instrumentation the contractor will be responsible for the installation, reading and reporting of results. Surface instrumen-tation will be provided and installed by the contrac-tor and monitored by an independent surveyor.
Due to the complexity of the project and the vari-ous new features in the design Sound Transit de-cided to extend the services of the SEM design team into the construction phase and thereby provide SEM inspection and support services.
Despite the fact that SEM is still a new technology the benefits are recognized throughout the industry. Owners realize that SEM can provide cost effective alternatives to other traditional methods. In some cases, where traditional methods would fail, SEM is the only viable method available to build tunnels.
Due to the need for infrastructure in many US cit-ies the utilization of SEM will gain more popularity and become a standard, cost effective construction method. With more projects the experience of own-ers, designers and contractors will increase and be-come more efficient and cost effective. It will and should be a method where new boundaries for tech-nical feasibility in tunneling can be established.
Some flexibility and adaptability is required when utilizing SEM the first time. This includes the evaluation of the contracting strategy, selection of experienced SEM designers, the utilization of the most economic excavation and support system and the use of a prequalification process if required.
During construction of SEM tunnels, especially in difficult ground conditions the designers should be involved and part of the SEM construction sup-port team. This is the only way that immediate re-sponses to construction problems are guaranteed and adjustments can be made efficiently.
The most advantageous contracting strategy has to be selected on a case by case basis. Experience shows however, that the traditional design-bid-build process is the preferable solution for complex, soft ground tunneling utilizing the sequential excavation method.