Large buildings in central London have traditionally been founded on piles bearing in London Clay. In the past decade and a half there has been substantial high value redevelopment in east London centering around the Isle of Dogs and Canary Wharf.
London Clay is largely absent in this area and Thanet Sand has replaced it as the favoured foundation stratum for bored piles. Designers and contractors alike have gained much experience in this stratum over the last 15 years and techniques have been developed which allow very efficient foundations to be installed in it.
Thanet Sand is relatively shallow at the Isle of Dogs (the surface of the stratum is 20-30m below ground level compared with 50-60m in central London). The accessibility of Thanet Sand at the Isle of Dogs and the fact that it is very dense has made it the preferred strata for founding bored piles.
Improvements in pile design procedures and piled-raft interaction (Chapman et al, 1999 and Nicholson et al, 2002) have led to higher loads being applied to these piles. Base grouted bored piles have been used to allow pile end-bearing capacity to be used at working loads. Typically, 40% of pile working capacity is derived from their end bearing.
The aim of the designer is to achieve an allowable load carrying capacity of large diameter bored piles (typically 1.5m diameter) which is limited by structural capacity of the concrete rather than geotechnical considerations.
It has been noted from pile tests that the pile base capacity decreases with increasing embedment into Thanet Sand. The loss in end-bearing capacity is partly offset by the increased pile length and shaft friction. However for larger diameter piles, an overall reduction in pile capacity with increasing embedment has been observed. To enable cost-effective design to be obtained it has been necessary to identify the zone of Thanet Sand that provides the highest overall capacity for the shortest shaft.
In the past, pile designs have used uncorrected SPT N values. The SPT blow count can be directly related to the end-bearing capacity (Meyerhof, 1976, proposed that for bored piles, the maximum end-bearing stress approximately equalled 120N in kPa and approximately 400N for driven piles) or used to provide an angle of friction, which in turn is related to a bearing capacity factor (eg Peck, Hanson and Thornburn, 1974).
The static cone penetration test (Dutch Cone) has been tried at Canary Wharf but has always met refusal before penetrating the Thanet Sand. It is therefore not a helpful design tool for bored piles in these ground conditions.
The SPT data from a recent site investigation in Thanet Sand at the northern end of the Isle of Dogs is shown in Figure 1. These have been undertaken and extrapolated using weak rock SPT procedures. The soil stratigraphy at this site is summarised in Table 1.
It can be seen that there is no appreciable trend of increasing or reducing SPT data with depth and that all the extrapolated blow counts are high (in excess of 200) and outside the usual range of values measured by the SPT test. For this data it would appear that end-bearing capacity would, at the very least, be constant with depth.
Figure 2 presents the corrected SPT blow count data (the correction for depth is based on Skempton, 1986). In this data there is also minimal trend for reducing blow count with depth (all data is above 50 blows) although there would appear to be less scatter with depth below 8m into the Thanet Sand. It should be noted that at other sites within 400m, SPT results of less than 100 have been recorded in the lower part of the Thanet Sand.
In addition to the SPT tests, a number of pressuremeter tests have been carried out at the same site using Menard equipment. This equipment can apply pressures of up to 8MPa and therefore applies representative stresses to the ground.
The Menard pressuremeter is not as sophisticated as the self-boring pressuremeter but has been used widely in continental Europe to provide design end-bearing capacities for piles (end-bearing pressure is directly related to the limit pressure, as described in Eurocode 7: Part 3).
The pressuremeter data is presented in Figure 3 and, unlike the SPT data, shows a very striking trend with depth. In the upper 7m of Thanet Sand the average limit pressure is in the order of 12MPa, while below 7m limit pressure has fallen to 7MPa.
This is a 41% reduction and would indicate a 41% reduction in pile end-bearing capacity for piles founded at depth. Similar reductions in limit pressures have been noted on sites where the SPT results reduce with embedment into Thanet Sand.
Possible reasons for the reduction in limit pressure may be variations in the grading and mineralogy of the stratum. Figure 4 shows the grading variation with depth. At the top of the stratum the average percentage of Thanet Sand comprised of silt and clay size particles is approximately 6% and this increases to about 15% at the base of the stratum. The trend is not striking and this is a function of the method of testing small samples (it has been found that using well mixed 'long samples' - ie sections of core - that are split before testing provides a better trend of increasing fines with depth.
This variation in percentage fines has also been assessed using geophysical testing. The results of resistivity profiling of Thanet Sand is shown in Figure 5 (higher resistivities are associated with lower clay contents and lower resistivities with higher clay contents). It can be seen that below about 8m into the stratum, resistivity is uniformly low (indicating a higher clay content) when compared to the material above 8m depth into the stratum. This profiling tool provides a very similar pattern of variation in the nature of the stratum as the pressuremeter shows the varying strength of the stratum.
A second possible reason for the reduction in pressuremeter limit pressure and hence a reduction in pile base capacity with depth is suggested in Figure 6. This shows the variation of hard (quartz and feldspar) and soft (mica) minerals with depth. Other sites have shown that the percentage of glauconite increases with depth. It is clear that there is a trend of reducing average mineral hardness with depth. The average Mohs hardness of the proportions of various minerals found at each depth has been assessed and is plotted on Figure 7. This confirms that the sand particles become softer with depth, although not perhaps by a great amount.
The increasing fines content and the proportion of soft minerals could have a large influence of the strength and compressibility of the Thanet Sand. At the high stresses associated with the base capacity of piles, the dilatancy of very dense sands is suppressed. High pressure triaxial tests are planned to investigate to investigate these effects.
At Canary Wharf a series of preliminary and contract pile tests have been carried out on base grouted bored piles founding in the Thanet Sand. The ultimate base stress (q m) at the pile toe level for the ground levels and water pressures appropriate at the time of the test. This ratio is called Nq*. Troughton and Platis (1989) discuss this ratio and reported a design value of Nq* = 47, based on tests on bored piles founded about 2m into Thanet Sand.
The results from recent pile tests are plotted in Figure 8. It is considered that greater reliance can be placed on the extrapolated ultimate base stresses for piles that settled more than 25mm during testing.The trend for reducing Nq* with embedment depth and into the Thanet Sand can be seen. At less than 5m embedment, the Nq* exceeds 60, while below 9m the Nq* is less than 35.This trend is similar to that consistently observed in the pressuremeter tests, but not always seen in the SPT results.It is concluded that the choice of ground investigation techniques used in Thanet Sand needs to be made carefully. Low technology tests, such as the standard penetration test, may not provide the level of information on the variability of the material required to reasonably predict piled foundation capacity and variation in capacity with depth.
However, tests which are more representative of the pile service conditions, such as the Menard pressuremeter, provide good quality zoning of Thanet Sand for use in pile design. Use of profiling geophysical tools can also provide a quick method for investigating the variability of the stratum with depth especially when compared with sole use of laboratory index tests (though 'long sample' grading tests also provide useful data) and provides a check on the variation in pressuremeter test data.
The Eurocode 7 design method for prediction of pile base capacity using pressuremeter test data does not provide guidance on the ratio between limit pressure and end-bearing capacity for bored base-grouted piles (data is currently only available for bored non-base grouted piles and driven piles). It is, however, considered that a bored-base grouted pile is likely to have load-deflection behaviour intermediate between a bored pile and a driven displacement pile. This relationship is under investigation.
Chapman TJP, Connolly M, Nicholson DP, Raison CA and Yeow HC (1999). Advances in understanding of base grouted pile performance in very dense sand, International symposium and exhibition in Tunnelling Construction and Piling, London.
Eurocode 7: Geotechnical design Part 3. Design assisted by field testing. ENV 1997-3 Fleming WGK, (1992). A new method for single pile settlement prediction and analysis. Géotechnique 42(3), 411-425.
Meyerhof GG (1976). Bearing capacity and settlement of piled foundations. Journal of Geotechnical Engineering Div. ASCE, 102 (GT3), 197228.
Nicholson DP, Morrison PRJ and Pillai AK (2002). Piled raft design for high rise buildings in east London, UK. Deep Foundations Institute, Nice 2002.
Peck RB, Hanson WE and Thornburn TH (1974). Foundation Engineering. John Wiley and Sons, New York.
Skempton AW, (1986). Standard penetration test procedures and the effects in sand of overburden pressure, relative density, particle size, ageing and overconsolidation. Géotechnique 36(2), 425-447.
Troughton VM and Platis A (1989). The effects of changes in effective stress on a base grouted pile in sand. Piling and deep foundations, London.
Yasufuku N, Ochiai H and Ohno S (2001). Pile end-bearing capacity of sand related to soil compressibility, soils and foundations, vol 41, no 4, pp59 - 71, Aug.