Instrumentation and Monitoring

The primary requirement was to monitor and record the detailed response of the building to induced settlement from the tunnelling works. The principal system of instrumentation used to obtain these observations was carefully selected arrays of horizontally and vertically aligned electrolevels supplemented by precise levelling (Figures 3.13 and 3.14).

The electro-levels, developed at the UK Building Research Establishment, had a range of ± 3° with a specified resolution of 1 arc second (Price et al., 1994). In practice, a repeatability of 2.5 arc seconds was reported, enabling the system to record slope changes to an accuracy in the order of 1/80,000. This was comfortably more than adequate to fulfil the task required. It should be noted that, although the capabilities of electro-levels had been in development during the 20 years preceding the project, the scale of their use for such a high-profile application of building instrumentation was innovative. Their performance is particularly sensitive to the installation procedures. These must be very precise. Even a seemingly minor inadequacy in how an electro-level is attached to its respective beam could, for example, make the readings unduly susceptible to temperature variations. Correct installation is one key aspect in

Figure 3.13 Primary instrumentation showing location of electro-level strings.

the procurement of instrumentation. Comprehensive advice on effective procurement is given by Dunnicliff and Powderham (2003). The readings were also most accurate within the middle third of their range. The OM team was keenly aware of such issues and the critical need for the monitoring system to provide reliable and consistently accurate readings. The success of the electrolevels at the Mansion House promptly heightened awareness of their potential and they were subsequently more frequently selected to monitor sensitive structures.

For the Mansion House a total 101 electro-levels were attached to the building, with fifty-five in the basement in four horizontal strings. Each of

Figure 3.14 West elevation showing height of ballroom and external electro-levels.

these beam-mounted electro-levels recorded changes in slope between adjacent reference pins set about 3 m apart. The other forty-six electro-levels were fixed individually in vertical lines to the external faces of seven of the principal masonry columns at the north end of the building (Figure 3.14).

Temperature-Induced Cyclic Movements

The performance of the system of electro-levels in real time proved so consistently accurate that it was possible to demonstrate that the temperature-induced

Mansion House 59

lateral movements caused in the building as the sun moved around it on a hot day were dimensionally more than the building settlement induced by the construction of the overrun tunnel. A plot of the variation in lateral deflection over a daily cycle caused by such temperature changes is shown in Figure 3.15.

These temperature-induced cyclic movements also highlighted the limitations of undertaking a conventional spatial survey with a manual theodolite because of the time required to complete a set of readings - especially since this involved setting up in the busy thoroughfares around the Mansion House. (In this context, to achieve the required accuracy, it shortly became commonplace after Mansion House to use remote reading total station theodolites. These overcome such daily cyclic variations by providing frequent and simultaneous multiple readings.)

Subsurface instrumentation also included horizontal and vertical strings of electro-levels installed in inclinometer tubes. These were to monitor ground movements and provide a useful correlation with any tunnclling- induccd effects detected in the building. Secondary systems included a water-levelling system and spatial surveys. The level survey points external to the building are shown in Figure 3.16. These had to be correlated with the internal reference points on the electro-level strings within the basement of the building.

Further details of the instrumentation, its performance and interpretation of the readings are reported by Forbes et al. (1994) and Price et al. (1994).

Figure 3.15 Diurnal cyclic movement measured by vertical electro-level string on west elevation.

Figure 3.16 Location plan showing external level survey points (e.g. MHI and 12/9).

Development of Progressive Modification

The implementation of the OM was based on the first overt application through progressive modification. With its basis set in defined construction stages, progressive modification enables progress from an established position of safety to a new one based on demonstrably robust predictions. As explained in Chapter 2, such a process had developed naturally in the application of the OM to cut and cover tunnels. For the Mansion House it was the cumulative effect of tunnelling beneath the building that was being progressively modified and evaluated through staged construction as described below. The procedure was as follows:

• (1) An assessment of the building and its foundation conditions was undertaken. This included a thorough condition survey and a comprehensive review of historical records.
• (2) Detailed consultations were made with the main contractor to review tunnelling methods and performance. Particular attention was given to the sequence of tunnelling and the level of risk relating to each of the remaining four stages.

• (3) The potential for tunnel realignment was also reviewed. There was no potential for this for the deep DLR tunnels and so was limited to the Central Line passenger link. This was at a relatively shallow depth and would thus not require moving far to be outside the contractual zone of influence (Figure 3.7). This left only the deep DLR tunnels for the over-run and the step-plate junction within the contractual zone of influence.
• (4) The courses of action planned in advance for significant deviations from anticipated behaviour were carefully assessed and developed in close coordination with the contractor and approved by the building owner.
• (5) The overall process was set out in a flow chart as shown in Figure 3.9. This chart set out the iterative process in which the effects of each successive stage on the building would be carefully monitored and assessed.
• (6) Approval to proceed to the next stage was dependent on the results of the preceding stage.
• (7) The overrun tunnel was undertaken first.

Critical Observations, Trigger Levels and Contingencies

The three zones of green, amber and red, as shown in Figure 3.8, set the basis for the ‘traffic-light’ approach. As noted in Section 3.3.3, this enabled the level of risk to be focussed on one critical factor - that of angular distortion. This had to be carefully assessed in conjunction with the ongoing condition surveys:

• (1) Trigger levels were set to initiate specific contingency measures. These related to the boundaries marking negligible, very slight and slight risk of damage to the building.
• (2) These boundaries were obtained by consideration of both angular distortion and horizontal tensile strain. The latter was relevant because the building would be subjected to a hogging deformation on the limb of the settlement trough.
• (3) With the risk level taking account of both effects, as shown in Figure 3.8, it was considered necessary only, and indeed practicable, to directly monitor the angular distortion.
• (4) Thus defined zones of levels of risk were established with trigger levels for angular distortions of 1/2,000 and 1/1,000 setting the boundaries as described below.
• (5) Measurements within first zone up to the first trigger level of 1/2,000 meant that tunnelling could proceed according to programme - providing no untoward adverse effects were evident from the condition surveys of the building.

• (6) The second zone between the trigger levels of 1/2,000 and 1/1,000 initiated the contingent action of higher frequency of reporting and additional condition surveys. Again, with satisfactory observations, approval would allow the next stage of tunnelling to proceed in sequence.
• (7) The third zone beyond a deformation of 1/1,000 meant suspension of the next stage of tunnelling, pending a comprehensive assessment of the effects on the building. If this proved significantly adverse, the options of installing one or more from the range of preventive works, listed in Section 3.2.2, would be evaluated.

The avoidance of such contingency measures involving major structural intervention to the building was of course one of the key objectives of the implementation of the OM through progressive modification. One of the more preferred options was structural strengthening in the form of steel ties. The locations and the ability to install them within a reasonably short period were carefully assessed in advance on site. Steel ties had been installed at the southern end of the building during Victorian times and were reported to have introduced some undesirable side effects by creating new zones of stiffness leading to differential movements. Such effects are analogous to the risks associated with partial underpinning. Local strengthening may act in opposition to the inherent rhythms of a building such as those arising from temperature- induced effects. If some form of physical intervention was deemed necessary, during the implementation of the OM, the viability of any such measure would be assessed in the light of the observed response of the building up that stage. Thus, it would have been on a far more informed basis than that which originally pertained before any tunnelling works. The existing instrumentation would have then been used to assist in monitoring the effects of its installation and subsequent performance.