As part of London's underground rail extension, corrective compensation grouting was applied to the Clock Tower housing Big Ben, England's most symbolic structure. The work was part of the massive Jubilee Line Extension Project, where vertically stacked 4.85-m-diameter running tunnels passed through Westminster with a centre line 28 m from the edge of the tower's foundation. In addition, for 160 m of the tunnel, excavation increased to 7.4 m in diameter to accommodate passenger platforms at Westminster Station. Also of critical importance, an 85 m x 65 m x 39 m deep surface excavation was cut just 34 m north of the tower to enable construction of access and station facilities for the existing District and Circle Line and new underlying Jubilee Line (Figure 1).

Having assessed the worst credible ground loss from these excavations, the associated settlement troughs, and structures at risk, a mandatory area of 50,000 m? had to be protected. This was accomplished by grouting through 35,000 m of tubes-a-manchette (sleeve port pipes) beneath more than 60 prestigious buildings, including the Palace of West- minster, the Treasury and the Institution of Civil Engineers, as well as a 3-m-diameter main sewer, a 1.22-m-diameter gas main, and a 0.76-m-diameter water main, all founded in the Thames Terrace Gravels overlying London Clay.

Concurrent and corrective compensation grouting involves the introduction of medium-to-high viscosity particulate suspensions into the ground between surface and sub-surface excavations and existing structures, in order to reduce ground settlement and structural distortion caused by excavation and tunnelling. To achieve this objective, appropriate grout is injected at high pressures to displace the ground by localised hydrofracture, but in a series of small injection increments through sleeve ports to control the propagation of the fracture planes and limit grout travel. On this project, concurrent compensation grouting was specified to minimise the effects of ground settlement during tunnelling. Corrective compensation grouting is reactive and was specified as a precautionary measure for the Clock Tower; thus, grout injection was only triggered when a pre-determined increase in tilt of the tower was observed. 

Above ground level, the 92-m-high tower comprises a 61-m-high by 12-m-square stone-faced load-bearing brick- work structure on which sits a 31-m-high cast iron belfry and spire. The tower weighs 8,540 tonnes and is founded on a 15-m-square, 1.2-m-thick rubble concrete raft resting on 3 m of Thames Terrace Gravels overlying London Clay. The tower had an historical tilt of 1/258, equivalent to a north- westerly lean (lateral deflection) of 213 mm at a height of 55 m.

Preparations for Grouting

From a temporary vertical shaft 4.5 m in diameter, 16 holes (127 mm in diameter) were drilled horizontally through the clay to create a fan array 7 m below the full footprint of the tower's foundation. The orientation and inclination of selected holes once drilled were checked using a borehole alignment device. A steel sleeve port pipe was placed in each hole and bonded to the surrounding clay with a low strength sleeve grout (water/cement (w/c| ratio = 2 by weight and 5 percent sodium bentonite by weight of water) produced in a high speed, high shear mixer/pump unit. The steel pipes, 68 mm in diameter with a wall thickness of 4 mm, were up to 65 m long with a maximum horizontal spacing of 2.4 m at the distal end from the shaft.

Rubber sleeves 150 mm long were spaced at 1 m centres, each covering two 25-mm-diameter ports as exits for the grout injections. Steel pipes were essential as repeated injections at high pressure would have led to deformation of traditional plastic pipe and subsequent jamming of the inflatable in-hole packer. Given the geometry of the array, the injection area was approximately 30 m long East-West × 24 m wide North-South, with only 6 m of width under the northern part of the tower's foundation (Figure 2). This injection area was divided into 3 m square blocks within which one sleeve was selected for an individual injection during each phase of grouting.

Traditional high fluidity cement/bentonite/grouts, e.g. W/C= 2-4 with 4-6 percent bentonite, are easy to pump but during hydrofracture of London Clay typically create fractures only 0.5-1 mm thick. To create thicker (10-25 mm) grout fractures over shorter travel distances using relatively low injection pressures, a wide range of lean economical flyash/cement/bentonite grout pastes was developed at the University of Bradford and designed to be suitable for low speed mixing and injection through small diameter sleeve ports. The grout pastes had pulverized fuel ash (PFA)/Ordinary Portland Cement (OPC) ratios of 9-22 with 4-6 percent bentonite and w/c ratios of 5-10.

Prior to use, all grout pastes were tested for vane strength, slump using an ASTM C143 cone (8" x 4" x 12" high), pressure filtration, and compressive strength development, followed by mixing and pumping trials. This testing resulted in an optimal low-cost grout design of PFA/OPC ratio = 20, 6 percent bentonite and w/c = 10 with a low 28-day unconfined compressive strength of 0.4-1.0N/mm? to facilitate re- injections. In practice, the water content was adjusted to give a slump of 180-220 mm that provided the required tooth- paste consistency and could be mixed and pumped readily using a mortar mixer/pump unit. During field injection trials to expose grout-filled fractures, 50- and 100-litre injection episodes created 20-mm- and 30-mm-thick wedges just beyond the sleeve ports with spreads of 1 m and 2 m respectively, illustrating appropriate fracture thickness and good control of grout travel.

In advance of corrective compensation grouting, a pre- conditioning phase involving 20 litre individual injections through alternate sleeve ports, at 2 m centres along each sleeve port pipe, was conducted. Sufficient grout was injected to stiffen the ground such that when subsequent excavations developed ground movement, injection of additional grout caused ground heave.

Control of Tower Deflections

The locations and timing of grout injections were deter- mined by observed tilt deflections of the tower, the horizontal deflections being recorded at 30-minute intervals using an electronic plum line at a height of 55 m. A 1/2500 in- crease in tilt (deflection = 22 mm) was set to trigger corrective compensation grouting, although tower deflections were known to fluctuate up to 5 mm due to temperature effects. A 1/2000 increase in tilt (deflection = 27.5 mm) represented the maximum permissible limit for the tower during the excavations.

Figure 3 illustrates the effect of initial grouting of 135 litres per sleeve port carried out in three stages (primary, secondary-repeat primary sequence) and amounting to a total grout volume of 6.06 m3 over a limited injection area of 24 m x 12 m (21 litres/m?). The tower was tilted back 5.8 mm over a period of four days. Injection rates per stage varied between 10 and 20 litres/min at pressures of 25-35 bars. Subsequently, the grout injection volumes were varied (range = 1.40 m* to 4.57 m*) over a six-week period. During this experimental period, it was established that deflections could be controlled to within 2 mm and reduced 25-litre injections through individual sleeve ports on a weekly basis could maintain the tilt of the tower.

Grout quality was checked via a slump test for each batch, combined with specific gravity, bleed, and flow cone tests every three hours. In addition, 100 mm cubes were taken for seven-day compressive strength determination for each stage of grouting. During grouting, grout flow rate, pressure, and volume were monitored and recorded in real time during each injection through individual sleeve ports. Electronic data capture and real- time analysis enabled engineers to observe the effect of individual grout injections and to assess the optimal lo- cations and grout quantities for subsequent injections, and thereby control the development of ground heave and associated tower deflection.

Generally, as surface and tunnel excavations progressed, the tower suffered a gradually increasing tilt but was tilted back when required in 1.3 mm to 6.5 mm increments using grout quantities ranging from 1.4 m3 to 9.6 m3 per phase. However, the final large injection involved 13.33 m3 of grout to tilt the tower back 11.7 mm to accommodate post-excavation, time-dependent ground settlement due to consolidation. Overall, a total grout volume of 115 m° (equivalent to 215 litres/m? over the final injection area of 30 m x 24 m) was injected in 24 phases during a period of 21 months and achieved a total heave at the North-East corner of the tower of some 40 mm and a cumulative tilt reversal of 110 mm. Generally, in simple terms, each 1 m3 of carefully placed grout created a tilt reversal of 1 mm.

Conclusions

Ground and structural movements, even for an iconic structure such as the Big Ben Clock Tower, can be control- led by corrective compensation grouting to within a few millimetres using appropriate sleeve port pipe layouts and injection of low mobility grouts. Great care is required in the design and implementation of compensation grouting and reliance must be placed on the observational method and detailed real-time monitoring of grout injections, ground movements, and structural deflections.

Stuart Littlejohn, DSc, FREng, FICE, FIStructE, is Emeritus Professor of Civil Engineering at the University of Bradford in England. As a geotechnical engineering specialist, he acts as an independent con- sultant and was formerly Technical Direc- tor of the Colcrete Group of Companies, now part of the Keller Group. He can be reached at gslittlejohn@ntlworld.com