This special meeting, chaired by Professor John Burland of Imperial College, gave geotechnical engineers the opportunity to hear Professor Ralph Peck describe the development of the observational method on the Chicago subway project in the early 1940s. Peck was followed by Alan Powderham of Mott MacDonald, who gave an overview of four recent applications of the observational method in the UK, and Duncan Nicholson of Ove Arup & Partners, who described the new CIRIA guide to the observational method.
Peck last lectured in the UK in 1969 when he presented the ninth Rankine lecture, 'Advantages and disadvantages of the observational method in applied soil mechanics' Peck returned to this theme with a detailed description of the development of the observational method in his work with Professor Karl Terzaghi in Chicago.
Terzaghi's return to America in 1938 coincided with the start of work on the Chicago subway system. He became involved with the project after speaking at an engineering meeting in the city on the dangers of tunnelling in soft clay.Such was his influence, he soon found his consultancy services in demand. Terzaghi decided to work for the City of Chicago but only after securing the establishment of a soils laboratory run under the supervision of someone chosen by himself and a then unheard of fee of $100 per day. The man Terzaghi appointed to run the laboratory was Peck.
Very little information was available on the effect of large diameter tunnelling in soft clays on surface structures. No reliable measurements were available on the magnitude and distribution of the pressures on linings. The subway project provided an ideal opportunity to investigate these aspects.
The subway was primarily constructed in the Chicago Clay, which was found to be highly variable and particularly soft in the downtown area. The clay surface was 3m above the tunnel crown, with relatively pervious fill materials above. To determine the soil characteristics and to develop a geotechnical profile, continuous 50mm diameter clay cores were obtained at 100m intervals, down to 3m below the tunnel invert. Water content was measured every 150mm in these cores and at least one unconfined compressive strength (UCS) test was carried out every metre. UCS generally varied between 50kN/m 2.In the downtown area however, the clay was much softer, with UCS of less than 20kN/m 2.Even in the 1940s, the UCS test was considered old-fashioned, but Terzaghi believed that ground deformations and resulting structure movements were related to the stiffness of the clay, which was in turn related to the UCS. Eventually more than 10,000 UCS tests were undertaken on the project.
Use of tunnel excavating machines had been found to be inefficient in Chicago because of the softness and stickiness of the clay. A large number of the sewers had been built using benching and hand excavation using clay knives. This method was initially adopted for the subway (Figure 1). The crown was supported by steel ribs which were set on footings in the lower bench.
At street level, vertical movements of up to 300mm were causing significant damage to buildings. In an attempt to limit the damage, screw jacks were used to level walls and floors. However, the contractor was not ready to admit that the ground movements were a direct consequence of the tunnelling works.
This issue sparked great interest from Terzaghi, who believed that if there were problems in the uptown section of tunnelling then the settlements downtown would be of a greater concern. He suggested investigating the movements of the clay around the tunnel using 'squeeze tests' where the tunnel and surface movements were measured over an excavation cycle lasting for 72 hours. Surface movement was measured by the levelling of the ground and structures and tunnel face movement was monitored by driving 2m to 3m long spear heads ahead of the tunnel.Lateral movements were monitored by measuring the convergence of the side walls.
Tunnel movements ranging from 25mm to 80mm were recorded during the squeeze tests. The subsurface and surface observations implied similar volumes of ground movement.The correlation of subsurface and surface movements led the contractor to direct his efforts to minimise these movements by the modification of the tunnelling process.
Initially this involved the use of 'Monkey Drifts' which were excavated 6m ahead of the face. The drifts were located at the edges of the haunch of the tunnel where an H-beam was placed before the lower section was excavated. It was intended to transfer the load from the ribs in the crown into the H-beam and in effect prestress the system. This significantly reduced the ground movements compared with the original method. Squeeze tests were again used to check the movements.
The board of consultants agreed that downtown, even with the best techniques, the risk of excessive damage to buildings was too great, so it was decided to change the construction method. Excavation using a shield was chosen and although not an unusual method, it had never been used in Chicago. Propelled by hydraulic jacks, the 8m diameter shield cut into the clay allowing material to be removed through the six chambers. Temporary lining was then placed inside the tail piece followed by 750mm of permanent reinforced concrete lining.
Observations clearly linked surface movement to tunnel movement and demonstrated that that two-thirds of the surface movement could be eliminated by introducing the Monkey Drift system. Peck and Terzaghi had shown that by experimenting with different construction procedures an optimum design and construction process could be achieved. Peck stressed that no theory had been used and all deductions were made from observation.
A similar observational approach was used to assess the design of the permanent linings. Initially these were very heavily reinforced but overdesign had not been demonstrated because of the difficulty in measuring earth pressures. Terzaghi proposed that two full-size test sections be built; one in the hand-mined section and the other in the shield section, 6m and 18m long respectively (Figure 2).At both trial sections the heavily reinforced lining was replaced with a relatively flexible 150mm steel rib lining, and total loads and movement were measured using pressure cells and strain gauges. Based on these observations Terzaghi was able to make recommendations for lining design.
Peck emphasised that Terzaghi did not make any recommendations using theory alone, but rather judgement combined with a conceptual model and critical observations. The key to the success of the Chicago subway project was having a clear ground model and not how it was designed, but how the critical movements were controlled.
He stressed that once a clear picture of how the ground movements were related to the construction was available, it was possible to reduce movement by refining the construction process. The heart of the observational method was critical measurement and keen observation before looking at the theoretical implications. It was not a case of using theory to dictate design.
Duncan Nicholson presented a review of CIRIA Report 185, The observational method in ground engineering: principles and applications, which was published last year.
The report covers definition and history, concepts, technical considerations, management considerations, contractual framework, applications and conclusions and the way forward.
CIRIA defines the observational method as 'a continuous, managed, integrated process of design, construction control, monitoring and review which enables previously defined modifications to be incorporated during or after construction as appropriate. All these aspects have to be demonstrably robust. The objective is to achieve overall economy without compromising safety.'
The guide focuses on process (Figure 3).The need to improve the link between the core of the observational method process in the lower part of the diagram and the policy and organisational aspects in the upper part of the diagram through value management was emphasised.
Chapter 3 covers the concepts involved in the method. The following concepts were discussed: Dealing with uncertainty The observational method can deal with uncertainty in the geology, parameters and modelling and ground treatment. These are the three main areas the report examines. Examples were given from jobs Ove Arup & Partners was involved in, including driven pile length variation in weathered chalk in the UK, progressive modification from two temporary props to one at the Channel Tunnel Rail Link cut and cover section in Ashford and groundwater control in response to measurements for a dry dock in the Philippines.
Predefined versus observational method design The main differences are listed in Tab le 1.It was noted that, by addressing buildability, the CDM regulations lead the designer towards the concepts of the observational method.
Accordance with design codes The challenge for the CIRIA report was trying to move Peck's approach, which was based on most probable and most unfavourable conditions, into modern codes. It is recommended that most probable and moderately conservative conditions are used for servicability limit state predictions and most unfavourable conditions are used for ultimate limit state predictions.
Rapid deterioration It was stressed that this is an important area. Development of movements towards the limit state must be slower than the frequency at which readings are taken.
Trigger criteria The report explains the traffic light conditions which may be used: Green = safe site condition Amber = decision stage Red = implement planned modifications Implementation of the observational method Two approaches are discussed in the report. One approach based on Peck (1969) starts with most probable conditions. The other approach, progressive modification, starts with characteristic conditions.
Chapter 4 of the report covers technical considerations and sets out a framework and rules for the industry to work around.
Chapter 5 deals with management considerations. The report lists the targets needed on a job for culture, strategy, competence and systems. Management issues are essential to the observational method because of the importance of construction control in counterbalancing the level of uncertainty.
Therefore the management system as well as the technical aspects must be robust. Use of the observational method complies with the Egan report, which identifies five key drivers for change, all of them necessary for an observational method job to succeed. The importance of the observational method from the conception of a project in achieving the best cost reduction potential was stressed.
Chapter 6 discusses contractual issues including types of contract, contractual difficulties using the observational method, contractual risks and strategies. Value engineering clauses and novation in traditional contracts are very useful. A novationtype contract is being successfully used for an Arup project at Fenchurch Street, London, involving demolition and deepening of existing basements.
Conclusions on what the CIRIA report has achieved are presented in Chapter 10.It has:
established a definition of the observational method process clarified uncertainty and risk management integrated the observational method into design codes and site control procedures provided technical, managerial and contractual guidelines for the team helped Egan cultural change in industry summarised applications/case histories.
Ways forward are also suggested in Chapter 10. Among these, in terms of concept and technical developments, geotechnical engineers need to communicate hazards and opportunities better to the design team. Risk management principles need to be implemented. There is a need to understand factors of safety better and be able to predict behaviour approaching ultimate limit state.
Ways forward on management issues include better interface with project management and increased understanding of costs versus savings and cost control. Contractual improvements include the need for appropriate contracts to deal with large cost risk. There is also a need to be more specific in geotechnical baseline reports in contracts.
The discussion covered the ways forward on the observational method, including a presentation by Jason Le Masurier who spoke about research at Bristol University on improving management of the observational method process. Ralph Peck's view was that there are dangers in over-structuring the observational method because this can dull the ability to observe and make judgements.
Duncan Nicholson emphasised the importance of formal systems for management of the monitoring and review process, together with the link back to design. The formal system is required to avoid abuse of the observational method and to enable it to be integrated into codes.