Heathrow Express Cofferdam (1994–1995)

Introduction

The Heathrow Express (HEX) Rail Link is an important new connection between Central London and Heathrow Airport. It provides a frequent, direct service between the airport terminals and London’s Paddington Station (Figure 5.1) with a journey time of fifteen minutes. The line runs 19 km along existing track before branching off to continue for 6 km below ground to the airport’s Central Terminal Area (СТА), where it involved substantial underground construction. In October, 1994, these works were in crisis following the major collapse of the large diameter station tunnels during their construction in the СТА. This case history describes how, through the single team approach, partnering, value and risk management and technical innovation were deployed to minimise the delay and cost overruns resulting from the collapse. The application of the observational method (OM) played a key role. The centrepiece of the recovery solution was the construction of a cofferdam measuring 60 m in diameter and 30 m in depth. Its design was specifically tailored to facilitate the implementation of the OM through progressive modification.

Key Aspects of Design and Construction

Overview

The original target date for the opening of the HEX was 1 December 1997. The bored tunnelling south of the M4 had commenced 3 years earlier. It included 3 km of running tunnels to the СТА, a series of station caverns underneath the СТА and a further length of running tunnel from the СТА to Terminal 4. A major collapse occurred when the station tunnels below the СТА, failed in October 1994 (HSE, 2000). The primary linings for these large diameter platform and concourse tunnels were then being constructed with sprayed concrete linings (SCL) (Figures 5.2-5.4). Tunnel invert is at a depth of around 30 m below ground surface. Fortunately,

Figure 5. / Location of cofferdam in СТА and HEX connection to mainline railway.

there was no loss of life or injuries but substantial damage to the works and adjacent structures occurred. At this stage, potential delay to the project resulting from the collapse was estimated to be about eighteen months. Fundamental to the recovery plan was the need to safely repair the damaged works and reconnect the tunnels and minimise delay to the opening date. The СТА cofferdam was the centrepiece in this strategy, and its successful construction enabled the line to be opened on 25 May 1998, recovering a full year of the lost programme (Thomas and Bone, 2000). An important early decision in the recovery was the formation of the Solutions Team, following from the single team approach developed by the owner BAA Airports. The Solutions Team members were selected from the main stakeholders of the project: BAA, main contractor, lead designer and the loss adjusters with their consultant.

Figure 5.2 Extent of tunnelling construction at time of collapse.

Geology - Conditions Prior and Post Collapse

A site investigation, to evaluate the changed conditions, was initiated immediately after the collapse. The ground conditions in the СТА, prior to collapse, were relatively uniform with approximately 6 m of Terrace Gravels overlying the London Clay which has a thickness of around 60 m at this location. The London Clay overlies the Lambeth Group, which generally consists of heavily over-consolidated clays and sands. This in turn overlies the Chalk which is present at a depth of approximately 90 m below ground level.

The recovery strategy required early establishment of a ground model. This involved an iterative approach and, as new information became available, it enabled development and refinement of the model. Original ground horizon levels were carefully assessed and compared with those post failure. The focus was the top of the London Clay which had originally been about 6 m below ground surface and the new levels were mainly assessed from a series of shallow boreholes. A series of deep boreholes were also carried out to assess the condition in and adjacent to the collapsed tunnels. Detailed core logging was used with the emphasis placed on visual descriptions. Investigation and design development were proceeded in parallel, and it was important not to create unnecessary delay with a prolonged programme of laboratory testing. Two sets of data for the top of the London Clay, pre- and post-collapse, were collated. To achieve the best estimate of the contours for these two London Clay horizons, the data were statistically evaluated through a process known as kriging (Clark, 1979). The difference between the two kriged surfaces was plotted as contours of settlement as a result of the collapse and is shown in Figure 5.3.

This process indicated that there were four localised areas of highly disturbed ground. In view of the large collapsed volume, approximately 6,000 cubic metres, and the subsequent amount of excavation required, it was considered that there was likely to be significant time-dependent softening initiated as a result of the collapse. The excavation of the cofferdam would also create a further reduction in stresses leading to a prediction for soil strengths much lower than for typical conditions in London Clay.

On the basis of the site investigation and predictive numerical analysis, four zones were assigned within the London Clay (Figures 5.3 and 5.4). Zone 1 was undisturbed intact London Clay, but this was considered to lie beyond the active wedge of soil on the outside of the

Figure 5.4 Cross sections showing predicted zones of disturbance.

cofferdam. So, the performance of the cofferdam would be principally influenced by Zones 2 to 4. Each of these was assigned two sets of bounding soil properties. The first represented moderately conservative (MC) parameters for the mass behaviour of that zone on the cofferdam as a whole. The subsidiary set was assessed as worst credible (WC) values, representing local influences that might have occurred where pockets of the most severely disturbed soil in that zone could result in adverse loadings on the cofferdam ring. These properties are summarised in Table 5.1.

Contractual Conditions

The New Engineering Contract (NEC) (ICE, 1993) for the project was specifically adopted by BAA project to facilitate a less confrontational approach to construction. The NEC emphasises the principles of trust and co-operation within a contractual framework. The report,

Table 5.1 London Clay soil parameters at Heathrow cofferdam site.

Note: TD = tunnel datum (= ordnance datum+100 m), d = depth belowground level, MC = moderately conservative, WC = worst credible. Zone 4 extends to +95 TD and Zone 3 extends to +93 TD. Below +93 TD, Zone 2 MC apply.

‘Constructing the Team’ (Latham, 1994), had set out the parameters for enhanced performance to promote a healthier UK construction industry through teamwork and undertaking projects ‘in a spirit of mutual trust and co-operation, trading fairly and nurturing the supply chain’. This was followed by the report ‘Rethinking Construction’ (The Construction Task Force, 1998). Chaired by Sir John Egan, it provided the stimulus for the initiatives of the Construction Best Practice Programme and Movement for Innovation. The key emphasis on teamwork and creating a ‘win-win’ approach also led to the process of partnering (Brook, 1997):

“Partnering has emerged in a number of forms, partly to reverse the suicidal fall into institutionalised conflict with appalling relationships between contracting parties in the construction industry, and more recently as a means of securing more work by creating a competitive advantage”.

The Cofferdam

The shape eventually adopted for the cofferdam was circular, 60 m in diameter and 30 m deep. As discussed in Section 5.4.2, it provided a very optimal creation of space compared to other options such as a square layout (Figure 5.5). It utilised 182 large diameter stepped secant/contiguous piles for the outer wall and 255 large diameter bored piles for the base slab. The design and construction had to deal with disturbed and unstable ground (including gravel beds and London Clay), water filled voids and major subsurface obstructions. These included mass and reinforced concrete and large items of buried construction plant. Being right in the middle of the СТА, there were also severe spatial limitations and environmental issues to address. The application of the OM through progressive modification was key to the overall design and construction strategy to manage risk and maximise opportunities to recover time (Powderham, 1998). Key issues were the control of ground movements and the associated effects on the cofferdam itself and protection of adjacent structures while achieving fast track design and construction.

Figure 5.5 Square cofferdam option (not implemented) compared to circular form.