Observations, Trigger Levels and Contingencies

The critical observations were the measured wall movements. The construction method, sequence and cycle time were selected so that there was sufficient flexibility to allow for the taking and assessment of these measurements.

A traffic light system, following its original development at Mansion House, was established setting green, amber and red ranges for wall

Figure 4.6 Ropemakers Fields: cross section and prop layout.

Figure 4.7 Dundee Wharf: cross section and multilevel prop arrangements.

movements. Trigger levels were accordingly set which related both to the amount and rate of movement (Figure 4.9). The first trigger level of 20 mm set the upper boundary for the green zone. Movements contained within this zone were not a cause of concern. The second trigger level of 25 mm marked the upper boundary for the amber zone. Measurements recorded in this zone would signify that wall movements were a potential cause for concern. This would initiate the contingency of increased frequency of monitoring. The focus here would be on assessing whether the rate of wall movements would mean exceeding the second trigger level before the blinding strut was cast at formation level. Such tracking would allow early identification of any adverse trends and so enable the timely implementation of contingency measures. If measured wall movements were consistently less than the first trigger level, then the design would be modified by initiating a construction sequence without midheight props. Also, if consistently low wall movements were sustained,

Figure 4.8 Initial construction sequence.

there would be potential for the additional progressive modification of increasing the length of the excavation bay and/or reducing the thickness of the blinding strut.

The contingencies initiated for wall movements entering or predicted to enter the red zone were as follows:

• (1) If it were during the soft prop stage - that is, with a gap between the prop and the wall, the use of hard props would be resumed and possibilities for a reduced amount of temporary propping only would be assessed.
• (2) If it were during construction without props, installation of the contingency prop would be triggered. Modification of the design would then have involved assessments to reintroduce more temporary support. This could have included, for example, shorter excavation bays or the reintroduction of hard props but at greater centres.

Measured Performance

The observed wall movements during the first propping trial strongly indicated that construction could generally proceed without the mid-height props. However, concerns regarding the actual performance, given the variable range of soils, placed constraints against proceeding on this basis. Thus, as noted, each construction front commenced with a propping trial before the viability of the OM design was confirmed by favourable observations and was accordingly approved and progressed.

The OM was introduced at the start of the first tunnel construction front in Ropcmakers Fields (Figure 4.6). The first stage at a construction front started with the installation of hard props in accordance with the original design (Figure 4.8). These were nominally pre-stressed to 10% of the specified design load. Prop loads and temperatures were monitored as excavation progressed, the former being measured using vibrating wire strain gauges linked to an automated data logger. Actual prop loads were assessed allowing for temperature corrections. The loads were found to be considerably lower than was originally assumed and typically not greater than the nominal pre-load (Table 4.2).

The second stage repeated the first except that the 100-mm-thick blinding at the base slab formation, specified in the original design, was increased to 300 mm. This was intended to act as a robust blinding strut. When the blinding gained sufficient strength, the props were dcstressed and wall closure was monitored. The third stage was to install ‘soft props’ for the propping trial. These had a 25 mm gap between the diaphragm wall at one end of each prop compared to hard props installed with no gaps. This allowed easy monitoring but would acceptably limit wall movements before the prop started to act. Up to this stage, the propping system as originally designed was still being used.

As the observed wall movements remained comfortably acceptable, the next stage was to proceed without props. Excavation started at the designated prop level and was taken down the full width of the tunnel to the base slab formation. This was limited to a length of 5 m and the blinding strut was cast the same day. Measurements of wall movement were taken at designated positions and times during construction. This was done using tape extensometers and surveying techniques - the former to monitor convergence and the latter to establish absolute movement. These measurements correlated well with those obtained by monitoring the gap in the soft props. To enable prompt action to arrest any adverse trends towards unacceptably large wall movements, a contingency prop and a number of reserve props were kept in the tunnel beneath the roof slab near to each construction front (Figure 4.8). (The overall process developed for observations and actions is described in more detail in Sections

4.4.6 and 4.4.7.)

Observations were taken at two-hourly intervals during the initial stages. When a satisfactory trend of wall convergence had been established, this frequency was reduced to twice and subsequently to just once a day. All the measurements of the wall movements were plotted and compared with a chart of predetermined zones and trigger levels (Figure 4.9). The allowable magnitude and the rate of movement which determined the boundary of each zone were set and agreed between all parties before any props were eliminated. A maximum wall convergence of 70 mm (35 mm for each wall) was allowed at formation level. However, in over 19,000 observations, the maximum convergence recorded was 11 mm and was generally less than 7 mm. Given these results, it is interesting to note that the diaphragm walls actually deflected more overall when the mid-height propping of the original design was used. This sounds counter-intuitive but was because the wall deflections were not effectively arrested until the base slab was constructed. With the midheight props in place, construction was much slower and during this extended period the walls still moved inwards below the props and of course compressive loads were also being generated in the props. So, when the props were finally removed, the walls moved inwards, thus adding to the displacement developed in the deflected profile during the time before the base slab was constructed.