| Project name: | Brentford Lock private housing development |
| Location: | Middlesex, UK |
| Keller company: | Keller Ground Engineering |
| Client: | Building contractor Taylor Woodrow Construction |
| Techniques used: | Basement construction using K-Wall |
| Date: | 2002 |
Keller’s novel and innovative watertight retaining wall system is based on slurry walls reinforced with closely spaced steel I-beams With most conventional retaining wall techniques there is trade-off between stiffness – needed to resist lateral ground movements - and water resistance. That makes steep-sided excavations in water bearing soils a particularly expensive geotechnical challenge. Keller’s K-Wall system provides an alternative, potentially low-cost watertight retaining solution, suitable at least for single-storey basement excavations. The system was developed by Keller’s US operation Hayward Baker, but the company recently completed its first UK K-Wall application on a housing development in Brentford, Middlesex. Main contractor Taylor Woodrow Construction is building a series of “basemented” apartment blocks and townhouses for developer Charles Church at Brentford Lock – an infilled piece of land between the Grand Union Canal and River Brent. The development means lowering the site to 2.6m below existing ground level and the adjacent waterways. Most of the excavation is in permeable made ground, but its base cuts into underlying soft to firm alluvial clay. This is underlain by sand and gravel, and so to achieve a cut-off between the rivers and the excavation necessitates penetrating the London Clay, to a depth of about 10m. The K-Wall is a slurry trench wall reinforced with steel I-sections. Where conditions allow its use, it is not only cheaper and quicker than a traditional secant piled or diaphragm retaining wall, but it also offers considerably improved watertightness. The system relies upon the arching effect, in which lateral loads bridge across adjacent I-sections. Clearly the spacing of the steel sections and the width of trench are critical to the design. Laterally, there has to be sufficient thickness of slurry to allow full arching to develop. The strength of the slurry wall mix then has to be able to withstand the compressive stresses this arching induces. However, as the strength increases the permeability of the mix must be maintained. Keller Ground Engineering used its data bank of over 5000 logged laboratory test results from previous slurry wall installations, to meet these exacting and unusual requirements. The design methodology is to carry out an ultimate limit state analysis to determine the depth of the wall, followed by a serviceability limit state analysis to determine bending moments, shear forces, deflections and net pressures on the wall. The design is completed with what Keller calls K-Wall analysis to determine the I-section size and spacing, together with the slurry strength. Prior to starting on site, Keller undertook a laboratory-testing programme in conjunction with Birmingham University on specially constructed scale model slurry wall panels. At Brentford Lock, Keller installed 550m of perimeter wall, 330m of which was reinforced with I-sections. The remainder had no retaining function, but was needed to provide a nominal hydraulic cut-off into the London Clay. Keller formed the 600mm wide slurry trench into the London Clay to depths of between 8.5m to 10.5m using a long reach back actor. Steel I-sections were placed at 1-1.5m centres together with a reinforcing mesh to reduce the risk involved in such a novel application. The project was Keller’s first use of the K-Wall system in the UK, which required detailed forward planning in determining potential construction risks and how best to carry out the work. But according to those on site is was “essentially like a slurry wall contract, except the mix was much stronger”. I-sections were then lowered in by crane using a guide template at ground level to maintain their vertical alignment. Following excavation of the basements, movement was actually less than the intentionally conservative design predictions. But back analysis of the wall’s performance will give Keller vital performance data that should lead to even more economic K-Wall designs and enable its use where retained heights are greater. |

