Report on the Denis Waite Lecture 'New developments in foundations', an informal BGA discussion held at the ICE on 14 November 2001, by Janice Windle and Vicky Potts, Arup Geotechnics.
Increased development in crowded urban environments has led to greater numbers of tall structures and deep basements, and these require deep foundations. The 2001 Denis Waite lecture on deep foundations was presented by three speakers who have been involved in the development of innovative schemes for such projects.
Duncan Nicholson of Arup Geotechnics described advances in the design and construction of base grouted piles used on phases two and three of the Canary Wharf development in London Docklands. The project involves construction of 12 office blocks, each up to 40 storeys, founded on large diameter piles. The piles will be founded in the dense Thanet Sand and are base grouted for extra stiffness. The water table at the site has been lowered but the piles have been designed with the long term water profile in mind.
Early on in the project it was found that the pile base capacity decreased with increasing embedment in the Thanet Sand.
Extensive site investigation involving borehole geophysics, shear wave velocity testing, pressuremeters, high pressure triaxial testing and mineralogy was undertaken to aid understanding of the Thanet Sand behaviour (Ground Engineering March 2002). The phenomenon is thought to be due to an increase in the amount of glauconite - a weak mica mineral - in the Thanet Sand with depth.
The design process has been improved using the results of the pressuremeter tests. These were back-analysed to provide more representative parameters for the design of piles, leading to increased loads being applied. Pile raft interaction has been modelled using 3D finite element analysis, which significantly reduced the number of piles needed.
Pile construction procedures included the dewatering of the Thanet Sand over an extensive area, a novel bucket design for cleaning the pile bases, sonic logging to detect pile defects and stringent control procedures for concrete quality, base grouting and the production of close-out reports.
Nicholson said developments arising from this project would include improved site investigation techniques for silty sands, improvements in the use of polymer muds, base cleaning and inspection tools, integrity testing, concrete stress levels and the understanding of the need for base grouting during construction.
Asked what software had been used in the pile raft analysis, Nicholson replied that the interaction was analysed using Dyna 3D, using a model with 75,000 elements and 0. 75M nodes. Abaqus was also used, less successfully.
Hugh St John of Geotechnical Consulting Group discussed the reuse of foundations. He said previously developed sites with basements and deep foundations are beginning to reach the end of their lifespan and be redeveloped.
Frequently these sites are very congested below ground, with structures built on under-reamed piles. Foundation removal can be expensive, but if previous foundations remain there is limited space for installing new ones.
Developers are increasingly finding no viable alternative but to reuse the foundations.
St John said that while this is not an advance as such and is not in the interest of piling contractors, designers will be forced to consider it as more large scale developments reach the end of their working lives.
Limited guidance is available to designers on this topic, as little has been published. Initiatives include talks that Arup Geotechnics is organising for the London District Surveyors Association and Building Control and an industrysupported BRE project.
It is important to ensure that none of the materials (pile, soil, or pile caps) are overstressed, and that foundations have acceptable load/settlement characteristics.
With old foundations it is known what they were intended to be, if records can be found and their existence can be ascertained by exposing the pile top. It can usually be assumed that the foundations have performed as intended.
At present, unless each pile is tested, the best that can be done is to limit the load reapplied to a pile to less than the previous maximum load. This causes difficulties where composite systems of new and old piles are designed, as the distribution of load between the new and old piles through the pile cap needs to be understood. It is also necessary to remember that old piles may be very stiff on reloading.
Asked about the reaction from building officers and the district surveyor, St John replied that for specific projects the approach had been positively received and the authorities were generally receptive to the idea.
The final speaker was Rab Fernie of Cementation Foundations Skanska who discussed Cementation's Cemloc system for remote placing of plunge columns. In top-down construction, columns are traditionally placed by installing a sacrificial casing, and sending operatives into the pile bore.
The Cemloc system positions the column from the top of the pile bore, from where it can be plunged into fresh concrete.
Rollers on the device can be adjusted to change the verticality and plan position of the loadbearing steel columns. Site operatives do not have to enter the pile bore, reducing health and safety risks, and sacrificial casing is not needed, saving costs.
Extensive research has been carried out, mostly to find the optimum shape of the plunge column cross section, and some empirical recommendations to facilitate safe bond stress design.
Discussion Professor Angus Skinner asked Nicholson if the test pile results shown for HQ3 were for the drawndown water levels. Nicholson replied that water levels were recharged from about -30mOD to -14mOD. Skinner asked for confirmation that the ratio of shaft friction to the working load was 0. 5 and that if this were the case, then the mobilised shaft friction shown seemed very low.
Nicholson replied that the test values were 1. 5 times greater than the design values.
Chris Raison of Chris Raison Associates asked about the angle of friction used in the design of the piles at Canary Wharf. He noted the high SPT values recorded in the Thanet Sand, but that the f9 value of 36° and Nq* values of 30-60 that were adopted for design seemed low.
Nicholson replied that triaxial testing showed that the f9 value changed significantly with changing horizontal effective stress. At low effective stresses, say 300kPa400kPa, the material dilates, whereas at higher compressive strengths the material compresses giving lower values of f9.
Nicholson believed this was due to the compressibility of the glauconite and silt particles in the Thanet Sand and that it would not be reflected in the SPT results.
SPT results alone were therefore not a reliable basis for design. He also noted that the design had to accommodate a rise in water level in the long term case.
Professor Richard Chandler, Imperial College, asked Nicholson what the purpose of base grouting was in this instance. Was it to grout the ground solid to increase the capacity of the Thanet Sands and the bulb of pressure at the pile base, or was it a form of compaction grouting, to fill cavities?
Nicholson replied that the primary reason for grouting was to try and eliminate any uncertainties in workmanship and unify the base performance, to allow uniform group load settlement. He also noted that he suspected that some grout also flowed up the side ofthe pile, increasing the shaft friction, but that it was not possible to know exactly where the grout had gone.
Rab Fernie asked Nicholson what happened when the friction was reversed. Would the strain gauges be able to detect this?
Nicholson replied that the small movements experienced on Canary Wharf, between 0. 2mm and 2mm, would not be sufficient to mobilise negative skin friction.
St John added that grout tubes were a relatively cheap method of quality control as they provided a means of sonic logging. Fernie added that although this might be the case, care was still required as a badly base grouted pile was worse than a normal pile. St John replied that this was not due to the base grouting but when and how the pile was built and the level of confidence it was possible to have in the contractor's ability to correctly form the pile base.
Geotechnical engineering consultant Arthur Penman said that before the introduction of base grouting, he advised on underreamed piles being installed in London Clay for a bridge over the River Swale on to the Isle of Sheppey. Problems were experienced, with pieces of clay falling from the roof of the under-ream.
The bridge was very sensitive to differential settlements so it was essential that the pile concrete went on to a clean floor.
Penman went down in a bucket to examine the problem and retrieved samples, a common practice at the time. This was not possible under modern health and safety guidelines.
Raison asked Nicholson if the shaft friction was taken as 0. 8 times effective horizontal stress, how sure was he that it was the ground stress controlling shaft friction and not concrete stress locked in during construction.
Nicholson did not know, but said the pile test was carried out with increasing water pressures which gave a range of effective horizontal stress values. If grout was flowing to the sides of the pile he could not state the effects of this grouting activity on the horizontal stress.
Angus Skinner asked the panel how, if piles were designed so conservatively, could they be assessed for re-use in the future.
St John replied that current methodology is to limit the new imposed loads on a pile to less than those previously applied.
The load ratio between new and old piles is still under investigation, as past performance is a critical issue.
Skinner asked how a composite system could be designed if engineers did not know how new piles would react with the old ones. St John answered that it was much more difficult to determine the stiffness of an old pile compared with new ones, and that research was needed into establishing the stiffness of old piles.
BGA secretary Neil Smith said that although it would be impractical to load test all piles surely some tests were better than none, and asked if any tests were being carried out. St John said it was practically and economically not sensible to test all piles and that clients were reluctant to pay for tests. He was not aware of any sites were old piles had been tested.
Professor Richard Jardine of Imperial College said he had done some testing of old piles both in France and in the UK. The results indicated that, over a six-month period, ageing led to a significant increase in the capacity of driven piles, but that this was not found to be the case for bored piles.
Dickie Bassett said that between 1966 and 1968 he worked on a project combining new and old piles but that this had resulted in differential settlements that could not be overcome. Slip units were developed, made from material from the Coal Industry Board, to 'cover' the old piles, allowing the new piles only to take the load and therefore settle (he referred to a 1968 paper by Clark PJ, Bassett RH & Bradshaw JB).
He continued by asking St John for his views on the possibility of overstressing the concrete in the piles and whether or not he felt that preloading the site before new pile construction could prevent differential settlements. St John felt the design of the pile cap was critical in distributing the stresses, and that it was therefore necessary for designers to look at the worst case scenario for the spreading of stiffness in the pile cap, and the range of stiffnesses required in the pile. He did not believe that this would be difficult to achieve.
If a site was full of piles, would it be possible to convert them to a piled raft, or could the piles be cut down to act as earth, another questioner asked. St John replied that either of these were possible and that any such ideas should be explored.
David Beadman of Arup Geotechnics asked St John if the durability of concrete in the pile was an issue. He replied that it was critical and referenced a paper by Chapman et al published in 2001 on the issue of durability and the importance of assessing site conditions which may induce sulphate attack. This made the choice of site investigation highly important.
He noted that excavation to access a pile may also alter the pile's exposure to sulphate and bring about an attack. As a general rule St John believed that if a pile had been in the ground with no deterioration for 30 years then this could be extrapolated to future life. Unless there had been problems in the past he could see no reason to anticipate problems in the future.
Hilary Skinner from BRE gave an overview of a DTI-funded project to identify barriers and solutions to promote reuse of foundations. The project started in April 2001 and will be completed by March 2004. Literature is being circulated and she called on industry to collaborate to achieve maximum benefit from the project.