Deep soil mixing is used extensively around the world and is becoming increasingly popular in the UK, says Bob Essler.
Deep soil mixing, once rarely seen in the UK, is becoming more common as consultants and clients recognise the benefits of this very versatile technique.
Keller Ground Engineering completed a project in Southend, Essex in 1995 and since then has been involved in an increasing number of contracts.
DSM can be carried out in two basic ways, depending on the method of mixing the ground with the injected grout.
Wet soil mixing uses a fluid grout pumped to the mixing head by conventional grouting techniques. The method is usually reserved for large rigs operating with large diameter tools (1m to 3m) and has normally been used in the UK for encapsulating contaminated ground, although in other countries it has been mainly used for ground improvement.
For example, in the last four years Hayward Baker, Keller's US arm, has carried out more than 250,000m 3of wet soil mixing for foundation support while in the UK, Keller has done about 50,000m 3. Dry soil mixing, on the other hand, is the mixing of ground with powdered grout, typically lime or lime cement mixes. The technique is completely different from the wet system and the comparison is made in Table 1.
Usually, dry soil mixing is more appropriate for very soft clays and peats, as the introduction of the dry binder is helped by the high moisture content in the ground. Wet soil mixing can be used in most soils but in clays can produce excess material due to the higher volume of grout introduced. Strengths can also be lower because water is added to the system.
Since late 2000, the Keller Group and Swedish contractor PEAB have jointly owned LCMarkteknik, a company specialising in dry soil mixing. The firms have completed many projects worldwide using the system.
Four UK contracts with about 70,000m 3of treatment have been carried out for embankment support, shaft break in and out, pipeline support and car parking.
Dry soil mixing has been used extensively in Scandinavia and the Far East, mainly in Japan.
Wet soil mixing was also developed in Japan and more recently in the US.
Since 1999, there have been two developments in standardisation of DSM:
lEuroSoilStab lEuropean Standard WG10.
EuroSoilStab (see page 34, this month) was an EU-funded project to develop a design guide for DSM in organic clays and peat. It has recently been published and covers the investigation, design and execution of the system.
European Standard WG10 is nearing finalisation and is part of a number of standards published or being developed by CEN/ TC288. TC288's remit is the 'standardisation of the execution of various methods for geotechnical works' (including testing, control and validation). WG10 is responsible for DSM. The working group developing the standard is multidisciplinary and includes an Austrian representative from the Keller Group.
DSM's underlying design philosophy is similar to other forms of ground improvement - to produce a stabilised soil that mechanically interacts with the surrounding unstabilised soil.
The applied load is partly carried by the columns and partly by the unstabilised soil between them.
Where the surrounding soil is exceptionally weak, no account is taken of the contribution from the soil and the columns can be designed as weak piles. When used for slope stabilisation or other solutions involving lateral loading, the design assumes a composite shear strength of the ground based on the relative areas of the columns and ground, ie:
Cu (overall) = Cu (column) x Column area + Cu (ground) x Ground area Typically, 600mm diameter columns can be designed to carry loads of 50kN to 70kN and larger diameters correspondingly higher loads. The design process normally consists of simple laboratory trials to determine soil reactivity to various binders. A comparison is then made with the extensive project database to determine field performance and hence establish the column design. For larger projects, a full field trial can be carried out which, although relatively costly, can ultimately pay for itself because a reduction in binder content can bring significant cost savings.
Wet soil mixing is usually validated by taking core samples or block samples as the strengths are generally too high for penetrative methods. Dry soil mixing can be validated either by standard CPT, a standard column penetration test (SCPT) using a blade or a pull out reaction test (PORT).
The difficulty with both CPT and SCPT is the tendency of the probe or blade to deviate out of the column. The PORT does not suffer this problem as the blade is installed at the bottom of the column, through the mixing stem, before soil mixing and is therefore constrained to travel up the centre of the column after construction. For both SCPT and PORT the force is correlated to the shear strength of the column by a factor usually 10 to 11.
The following case histories help illustrate the use of soil mixing methods.
In Charleston, South Carolina Blue Circle Cement wanted to build two new 50m diameter cement domes with design loadings of up to 550kPa. Ground conditions comprised 2m of made ground overlying about 8m to 10m of very soft to firm clays which in turn lay on a Marl bedrock.
Hayward Baker's solution was to build 785, 2m diameter columns to transfer the dome load to the Marl. In all, nearly 15,000t of cement was used and the project completed in less than two months.
In the UK, the Queen Elizabeth Dock project in Hull demonstrated the benefits of alternative design using dry soil mixing.
Work involved construction of about 450m of perimeter retaining bund on the River Humber foreshore for land reclamation.
The site is within the confines of the Queen Elizabeth Dock and the ultimate client is Associated British Ports (ABP). Birse Construction won the contract on the original design of conventional staged placement of fill with associated monitoring of resulting pore pressure build-up. The ground conditions are 4m to 6m of very soft silty clays overlying glacial till.
Keller Ground Engineering was approached by Birse in late 2000 to develop a soil mixing solution to improve the upper 2m of the very soft silts to help platform construction. Keller proposed a redesign by extending the soil mix solution to the glacial till and effectively removing the requirement to limit embankment construction due to pore water pressure build-up.
This had the potential to reduce the construction programme by two years. Birse and Keller made a joint presentation to ABP to change the construction process. ABP consented and the Department for Environment, Food and Rural Affairs approved the change. Keller began a site trial in July 2001.
This had to demonstrate both short-term and long-term performance because the only way the columns could be built was to effectively use the soil mix columns built on the previous tide as support for the 35T soil mix rig. Design of the load transfer platform required a minimum Cu in the columns of 45kPa at 6-12 hours, the time between tides.
This meant the 16 trial columns were built one afternoon in late July. Birse, its consultant Halcrow and Keller returned at 1am to carry out insitu testing.
Results demonstrated the shortterm strength gain had been achieved and, subsequently, that the long-term design Cu of 150200kPa was also reached.
The Keller design was adopted and a performance specification written into the contract between Birse and Keller. Main works began in early August 2001 and phase 1 was completed at the end of September. Phase 2 starts this month and is planned to finish by the end of July.
In phase 1, two LCMarkteknik rigs were used in double shifts.
They worked from each end of the perimeter bund installing up to 60 columns in the three- to four-hour tidal window. Insitu vane testing was carried out on test columns at the start of each shift to check the design column strength had been achieved and PORT testing was carried out on works columns to validate the overall design.
Bob Essler is associate director at Keller Ground Engineering.