The design and construction of filled building platforms (engineered ﬁll constructed over low or uneven terrain)

Geotechnical Engineering Volume 164 Issue GE2 The design and construction of filled building platforms McNicholl

Proceedings of the Institution of Civil Engineers Geotechnical Engineering 164 April 2011 Issue GE2 Pages 89–99 doi: 10.1680/geng.2011.164.2.89 Paper 1000021 Received 22/02/2010 Accepted 22/07/2010 Keywords: buildings, structure & design/embankments/land reclamation ICE Publishing: All rights reserved

The design and construction of filled building platforms Denis Patrick McNicholl, MSc, PhD, CEng, FICE, FIStructE, FHKIE Director, Denis McNicholl Technology Ltd, Alton, Staffordshire, UK

Building platforms are described, and their purposes are explained. Historically, construction standards for engineered fill were not good, but following development of new specifications in the UK, and after major failures in Hong Kong, good practices were established to improve the construction of building platforms. In the UK this work was strengthened by the Building Research Establishment, with valuable contributions from individual authors. The paper summarises the advice in the literature in the context of almost 50 years’ experience in the UK, central Europe and China, and sets out recommended procedures for constructing filled building platforms. Site practices in the UK and elsewhere do not always follow the established guidelines. Several examples of good practice and departures from best practice are provided, with some photographs. For legal and commercial reasons the cases are not identified.

1.

Introduction

A building platform is defined as the engineered fill that is constructed over low or uneven terrain, and which supports the floor and foundations of a building. The platform may include a capping layer, but generally excludes traditional floor construction. Figure 1 shows a cross-section through a typical building platform. Platform engineering has become important in developing topographically difficult sites, in the redevelopment of brownfield sites, and because of a trend towards larger buildings, particularly for supply chain and industrial purposes. Although the consequences of failures in platform engineering can be serious, evidence shows that platform design and construction do not always follow best practice. So, by reference to the literature, experience and case histories, a 10-step procedure has been developed for the design and construction of filled building platforms.

The narrative and case histories are drawn from professional experience spanning over almost 50 years, and gained in the UK, central Europe, Scandinavia and greater China, including Hong Kong SAR.

2.

Historical summary

In the 1950s and 1960s the UK government increased housing production significantly, and to meet the need for new sites in the industrial Midlands and North of England new development platforms were formed by cutting and filling former coal waste and steel waste tips. Some earthworks volumes exceeded 250 000 m3 , and were carried out without a full understanding of compaction theory and best practice. Contemporary specifications called for the fill to be ‘well compacted’ or ‘thoroughly consolidated’; little attention was given to layer thickness, compaction methods or testing. Some engineers introduced end-product specifications to site formation contracts, but this introduced difficulties in setting target performance measures with acceptable tolerances. Argu-

Capping layer

Engineered fill

Benching Geotechnical separation layer Granular starter layer Underdrainage

Figure 1. Section through a typical building platform

89

Geotechnical Engineering Volume 164 Issue GE2

The design and construction of filled building platforms McNicholl

ments about sampling ensued, and there were problems in the time required for testing when high volumes of fill were being placed quickly. Sometimes it was too late to remedy nonconforming fill layers that had been covered by thousands of cubic metres of succeeding fill.

engineers relaxed requirements on layer thickness and compaction methods while claiming to work within the SHW guidelines.

As the UK National Specification for Highway Works (SHW) evolved, engineers began to use the Series 600 earthworks specification for the design and construction of filled building platforms, as well as for highways. Excavated or imported filling material was broadly classified as suitable (acceptable), unsuitable (unacceptable) or marginal (material that could be improved and used beneficially). By referring to detailed material classification tables, engineers could follow a reliable method specification that described how to spread and compact a particular type of fill material. The SHW tables and appendices provided details of maximum and minimum moisture contents for each material, expressed as optimum moisture content plus or minus an operating tolerance. This allowed specifying engineers to set target moisture contents that were slightly wet of optimum, and which produced fill of high relative density with air voids of less than 5%. Engineers could select the roller weight and number of passes for a particular layer thickness. This use of a detailed method specification brought simplicity and clarity, and allowed contractors to proceed quickly with earthworks, minimising disputes about the attainment of end-product specifications. Moreover, as layer thickness is easy to monitor, and provided the correct roller was on site, it became easy to count the number of passes. The SHW was not followed exactly in every instance. In the early 1970s, in Warrington New Town, where contract periods were short and where there were difficulties in working the local glacial soils, a policy decision was taken to use only granular material (and occasionally pulverised fuel ash (PFA)) for platform construction. Surplus unsuitable (unacceptable) cohesive and organic materials were diverted to the construction of large noiseattenuation mounds (up to 7 m high), which can been seen today north of junction 21 of the M6, and around the approaches to the M6/M62 junction. In the backfill and redevelopment of some large open-cut coal mines, the consequences of compaction failures were viewed as serious, and although the UK Specification for Highway Works was used as a base document, some engineers reinforced the SHW method-related approach by the use of control testing. It became possible to carry out rapid assessments of density using nuclear density gauges (NDG), and these were sometimes used to control and verify compaction in projects specified on the basis of SHW. Although SHW continued to be a guiding document for the placement and compaction of fills, case histories had arisen in the 1980s and early 1990s where filling carried out notionally to SHW methods became unstable, possibly because supervising 90

Later schemes benefited from the Building Research Establishment (BRE) report entitled Building on Fill: Geotechnical Aspects (Charles, 1993), which brought together the results of research on many filled sites and, most usefully, provided a definition of engineered fill. This work was extended by Trenter and Charles (1996), who noted that most engineered fills were based on the Department for Transport Series 600 specification, and introduced a model specification for engineered fill for building purposes. Although the model specification was limited to overall depths of fill not exceeding 5 m and for low-rise buildings, it provided simple, direct advice, and was later incorporated into the second edition of the BRE report (Charles and Watts, 2001). Charles and Watts provided an explanation of a mechanism now known as ‘collapse compression’. Partly saturated fills that might have remained stable for several years may sometimes settle abruptly, with no externally applied increase in effective stress, as a result of increases in moisture content brought about either by vertical infiltration or by long-term rises in the water table. This mechanism may be inhibited if the fill is heavily compacted to at least 95% of maximum dry density (MDD), and in clay fills to a state where air voids are less than 5%. Events in Hong Kong in the late 1970s also suggested that building platforms built by end-tipping or with relaxed layer thicknesses, and without due control, were prone to rapid catastrophic failure. Personal experience between 1978 and 1984 of call-outs in Hong Kong during typhoons, and after failures in fill platforms, illustrates that filled material that is not compacted to better than 85% or 90% MDD can fail in a rapid and brittle fashion when inundated. Debris may liquefy, and travel paths can extend over several hundreds of metres in steep topography. Eyewitness accounts and other evidence pointed to the extreme speed of such events. One such event occurred on 25 August 1976, at Sau Mau Ping in Hong Kong. A 30 m high filled building platform failed catastrophically, claiming 18 lives and injuring 24 others. According to eyewitnesses, the failure mass travelled at ‘about 40 miles per hour’, and according to those involved in the clear-up, the failed mass extruded itself through the ground floor and between the cross-walls of an occupied building up to the height of the ground floor ceiling (Bates, personal communication, 1980, 2010). The failure occurred very near to the site of a similar June 1972 failure on the same development platform, which claimed 71 lives. The Hong Kong Government’s Independent Review Panel on Fill Slopes (Hong Kong Government, 1977) reported that the failure occurred as a result of water penetrating into the exposed slope face of the poorly compacted platform, leading to consequent loss of strength and ‘an almost instantaneous conversion of the slope into a mud avalanche.’ The report pointed to the



structural collapse of poorly compacted soils when undergoing deformation, and to how this can lead to liquefaction. Compaction was said to improve stability by increasing density and strength, and the reviewers quoted a standard density of at least 85% as being necessary to prevent structural collapse of a deformed inundated fill. It was said that saturated permeability can be reduced tenfold by increasing compaction standards from 80% to 95% standard density. The panel also said that although adequate specifications for the construction of fills had been available in Hong Kong since at least 1966, there was little evidence to suggest that these specifications had been applied to any significant extent.

ence has shown that the methods suggested by Godfrey and Clayton are preferred by clients and non-engineering professionals, since risk quotients can be expressed in clear and simple terms.

As a result of these fatalities and others, the Hong Kong government (now the government of the Hong Kong Special Administrative Region) set up a Geotechnical Engineering Office (GEO) at the end of 1977, which has carried out a great deal of research, and has produced a series of documents that guide practising engineers in the investigation and methodology of earthworks. Best practice has been defined in GEO’s Technical Guidance Note 7 (Government of Hong Kong SAR, 2004). Although this document focuses on the recompaction of existing fill slopes, it has several recommendations that are relevant to new platforms in Hong Kong and elsewhere. (In the Hong Kong context, building platforms are termed fill slopes because of the emphasis on preventing failures in the exposed fill slope faces.) In addition to advances in compaction methodology, work in the 1970s by Knill, which was refined by Price and Lumsden (personal communication, 1982) and later published by Knill (1982), drew attention to the difference between the material and mass characteristics of soils, including fill, and established a system for considering engineering structures in the context of their engineering geological environment. Burland (1987) presented a clear view of geotechnical engineering comprising three nodes of a triangle: establishing the ground profile, defining ground behaviour and modelling (physical and analytical) – all interlinked, and considered in the context of ‘well winnowed experience’. This arrangement has become known as the ‘Burland Triangle’. The work by Knill and Burland is important in understanding how to conceptualise and analyse filled building platforms, including placement of the fill slopes into an existing environment, and in considering how the old and new structures will interact. This is particularly important in assessing elements such as underdrainage and benching, and in assigning geotechnical characteristics to the filled mass of the platform. In order to make the construction process more reliable, project teams are now using risk registers such as those introduced by Godfrey (1995) and Clayton (2001). The guidelines in these documents have been found to be easily assimilated by multilingual project teams. They are sufficiently straightforward that they can be used throughout the construction team, and experi-

Since the 1990s, long-term settlement of finished floors has become important everywhere in very large logistics and manufacturing buildings and also, sometimes, to meet the long-term settlement limits for high bay racking and high-speed crane access to very narrow aisle storage. (Floor areas of 60 000– 120 000 m2 are becoming more common, especially in China, Romania and elsewhere in central Europe.) The requirement for reliable long-term floor performance – that is, acceptable longterm deflections – has focused attention on the need for highquality engineered fill, and this also has prompted the reengineering to higher standards of previously reclaimed sites that had been prepared without modern controls. Finally, the institutional funding and later resale of large buildings require that the overall construction process be recorded and transparent, demonstrating that the whole project, including the building platform, has been constructed with due care. This is sometimes difficult if the earthworks and building are carried out by different agencies. These two recent trends have also encouraged the use of method-related specifications that are controlled and verified by index testing.

3.

Do we need further guidance?

Reference to case histories listed in the Appendix demonstrates that many engineers adhere to good practice, but some do not. Indeed, the number of non-conforming cases identified within the last 10 years indicates that good practice, as demonstrated in the literature, is sometimes ignored, occasionally with unfortunate consequences. To guide engineers in following best practice, the relevant literature has been summarised and, making use of Burland’s terminology, the advice from the quoted sources has been winnowed against experience and reduced to a 10-step procedure for the design and construction of filled building platforms.

4.

A 10-step procedure for the design and construction of filled building platforms

The elements of each step are summarised in note form. 4.1

Studies

1. Carry out a desk study. Should include a walk-over survey, aerial photographs, assessment of existing topography and drainage regimes, site history and existing filled zones, culverts and water courses. May include a topographical survey and a flood risk assessment. 2. Understand the engineering geological model. Use sections and three-dimensional (3D) views to understand how the new platform fits into the existing terrain, and how the old and 91



new will interact. Understand drainage regimes and water catchment areas. Design a site investigation to confirm understanding of the engineering geological model, and to characterise the engineering materials. Visualise potential problem areas. Figure 2 is an outline model of a fill slope that failed during rainfall, partly because the model was imperfectly visualised and understood. The potential presence of underlying shear planes, mentioned in memoirs, was not identified. 3. Site investigation and report. Should include a realistic and economically justified scheme of drilling and probing to confirm and validate the model, and to investigate difficult areas. Laboratory testing should be sufficient to give preliminary characteristics for underlying terrain, cut areas and excavation characteristics, and possibly for filling source materials. A good report would also pre-figure the likely foundation and floor settlement, based on the available project details. Set up a risk register. 4. Characterise fill materials at source. May be carried out as a sub-study of the main site investigation, but is usually carried out when cut/fill drawings have been carried out, and source materials have been identified. Establish maximum and target dry densities, and also optimum moisture content. If materials are to be differently sourced, devise a method for integrating them into a reliable filled mass. If material properties are inadequate, but clay-based, consider use the of lime/cement stabilisation. At this stage some engineers set target indices such as California Bearing Ratio and modulus of subgrade reaction (K value). In addition to conventional considerations of plasticity and moisture content, grading is also important. Some materials, at each end of the grading curve, are difficult Infiltration assisted by shrinkage cracks

Rainfall

to compact. These include fine sands and silty sands with uniformity coefficients less than 4, and gap-graded gravels or coarse fill (see Figures 3 and 4). The former may have to be mixed with other materials, whereas coarse and gap-graded materials such as those in Figures 3 and 4 are almost impossible to compact in layers of reasonable thickness. There have been failures identified as being due to inadequate compaction in thick layers of coarse fill and instability in gap-graded rock fill arising out of point-to-point contacts and point-to-plane contacts (see Figure 5). In short, gap-graded coarse fills require extra processing by crushing and grading. Migration of fines is also a possibility in coarse fills (Figure 6), and some form of geotextile separation layer may be needed. At source approval stage it is useful to calibrate nuclear density gauges against results from sand replacement tests. This process should be repeated during the course of filling for large projects.

4.2

Design

5. Design the platform filling. Design the interface between cut and fill areas. Consider basal drainage, platform drainage and slope face drainage. These are almost always required. The fill platform identified in Figure 2 failed in part because no underdrainage was provided. Set target densities, and an acceptable range of moisture contents. Prepare a specification. A simple approach based on Trenter and Charles (1996) is often better than large complicated specifications, particularly in international projects. Consider whether or not a basal granular starter layer and /or a geotextile separation layer are needed at the base of the fill. Pore pressure build-up No under drainage Pre-existing shear planes

Fill

Infiltration

Fill exerts pressure Infiltration

Through-flow on interface Through-flow in head/soft upper layer

Glacial till

Potential for elevated pore pressure in lenses

Figure 2. Model of a modern fill slope that failed during rainfall

92

Potential lenses



Migration of fines

Figure 6. Showing migration of fines in gap graded fill

The use of very large boulders, sometimes called ‘plums’, may be appealing in difficult terrain (see examples in Figures 7 and 8; photographs taken in Scandinavia in 2009), but it is often a source of problems, because it is almost impossible to compact such material, and successive layers often become gap-graded through migration of fines. The placement of a selected intermediate layer above the boulder layers is very expensive, and the jagged nature of the boulders rules out a conventional geotextile separation layer. Check the settlement of platform and underlying soils under service loading. If settlement is excessive or unreliable, then consider ground treatment (including surcharge) to fill or underlying in situ soils. Prefigure the floor slab design, and whether or not a capping layer is required. Consider instrumentation. Refine risk register and price alternatives.

Figure 3. Gap grading

4.3

Figure 4. Unprocessed quarry material being spread and compacted by Hymac

Point-to-point contacts

Construct

6. Site preparation. Carry out excavation; proof-roll basal formation and remove soft spots; install underdrainage. Bench into the existing terrain as necessary, restricting bench thickness to 300 mm. Install instrumentation; install granular starter layer or other basal layer as required. Set up site lab,

Point-to-plane contacts

Figure 5. Showing point contacts in ungraded fill

Figure 7. Boulders to be placed in basal layer

93



Figure 10. Compaction in progress at a large site, Rugeley, Staffordshire, UK Figure 8. Placing boulders in basal layer

or procure visiting test facilities. Carry out compaction trials on large jobs and re-calibrate index testing, such as plate bearing test and NDG results. Arrange ventilated and secure storage facilities for samples and equipment. 7. Placement and compaction. Carry out filling works to predetermined methodology and complying with the specification. Place differently sourced material in layers, or as designed, in order to engineer the fill to a reliable endproduct. Control layer thickness (rarely should this exceed 300 mm); Figure 9 shows non-engineered fill laid in 2 m thick layers; validate roller type and roller weight. (Figure 10 shows compaction in progress on another site.) Ensure that layers are more or less horizontal, or regular. Provide temporary drainage facilities in poor or doubtful weather. 8. Control testing. Carry out or commission independent monitoring. Initiate control testing on each layer. Use plate bearing tests to assess surface modulus values, and as an index of performance. Additionally or alternatively, use a Figure 11. Nuclear density gauge

Figure 9. Non-engineered fill in Scandinavia

94

nuclear density gauge (NDG) (Figure 11) to assess whether or not target densities have been achieved. The usual frequency is one test per 1000–1500 m2 of floor area. Carry out occasional sand replacement or other tests to verify test procedures. Ensure that the results and locations of tests are properly recorded. 9. Check imported materials. Filled platforms sometimes require imported materials for granular starter layers, and for capping layers. It is important that delivery certificates are thoroughly checked, and that independent testing is carried out, as the quality of materials often varies day by day, and ‘off-theshelf’ suppliers’ certificates are sometimes erroneous. This is particularly the case when there is a large proportion of recycled product, or in overseas locations where national testing regimes are not well developed. Figure 12 shows non-



Figure 12. Non-conforming granular fill

Figure 14. Plate bearing tests in Zagreb on finished platform

Figure 13. Non-conforming capping layer

conforming granular fill, and Figure 13 shows nonconforming capping layer material. Both had conforming certificates. The checking of imported materials by direct observation is very important. 4.4

Verify

10. Verify the completed works. On completion of the platform, it should be proof-rolled before placing the capping layer. Carry out further verification testing on the completed formation. The usual frequency is one test per 500 m2 . Use plate bearing tests or similar to establish K values, CBR, and – in continental practice – Ev1 and Ev2 values. Figure 14 shows a plate bearing test in progress in Croatia, and Figure 15 shows

Figure 15. Heavy load plate bearing test Zagreb

a heavy load plate bearing test on the same site. For surety, consider carrying out DPT testing at the rate of one test per 3000m2 . If a DPT rig cannot be mobilised, then a simple sitefabricated penetrometer can be used to locate soft or loose areas. Figure 16 shows a site-manufactured probe that was used in Croatia. For institutionally funded buildings and some others it is usual to 95



Load ⬃4–5 kg

50 cm

Gas pipe Dia. ⫽ 30 mm

Sliding of rebar into pipe

Rebar Dia. ⫽ 20 mm

Figure 16. Site-made penetrometer

assemble a completion report, with details of all elements of the procedure, backed up by test data.

5.

Concluding remarks

(a) Although there is advice in the literature, some engineers fail to give adequate consideration to filled building platforms, and failures can and do occur. (b) Filled building platforms, if properly investigated, designed and constructed, can perform satisfactorily in service, and even deep fills of 30 m thickness or greater can be engineered to provide a competent long-term subgrade for floors and foundations. (c) A 10-step procedure has been outlined; it provides a check list and advice on how to approach filled platforms, and how to ensure that they are adequately constructed. (d ) The principle steps may be summarised as follows. (i) Carry out a proper desk study. (ii) Understand the engineering geological model and the placement of the filled platform in the existing terrain.

No. Location

Date

(iii) Validate the model, investigate the fill site in general, and directly investigate the underlying terrain and the filling source material. (iv) Characterise the fill materials, and take care in assessing very fine or very coarse, gap-graded materials; set target standards for dry density and moisture content; carry out index testing, and calibrate special testing equipment. (v) Design the filled platform, including the interface between the fill and the existing terrain and underdrainage, after choosing a specification, preferably modelled on Trenter and Charles (1996). Having regard especially for layer thickness and grading, go on to estimate the likely settlement under service loading; consider the need for ground treatment; and establish a risk register. (vi) Carry out site preparation, including engineering the interface between fill and the existing terrain, underdrainage and instrumentation; then carry out compaction trials and/or recalibrate test instruments. (vii) Carry out the filling works according to the specification, paying particular attention to layer thickness and grading. (viii) Carry out control testing to a predetermined schedule. (ix) Check imported materials, relying on observation and not wholly on conforming test certificates. (x) Verify the works by carrying out testing to a predetermined schedule. (e) Caveat aedificans (‘Builder beware’). Based on experience gained of call-outs during Hong Kong typhoons between 1978 and 1984, it is important to note that filled platforms constructed of medium-density materials (say 80–85% MDD) can fail catastrophically if stressed when saturated. The failed mass can move very quickly – almost explosively – and this may not allow people to evade the failed mass. The construction of filled building platforms therefore requires due care and respect.

Appendix. Selected case histories For commercial and legal reasons these case histories are abbreviated, and presented in note form.

Building end-use Special features

Related to studies and design 1 UK Midlands 2002/03 Storage and distribution

7–12 m high embankment on glacial site with pre-existing shear planes mentioned in geological memoirs but not highlighted in desk study or identified in trial pits. No underdrainage to embankment. No platform drainage (Figure 2)

Outcome

Slope failure after rainfall

( continued) 96


No. Location


Date


No specific desk study; no project-specific site 2008/09 30 000 m2 storage, packing investigation. Building situated over infilled lagoons and distribution with no specific ground treatment Related to design and construction

Outcome Under monitoring

2

UK

3

Leabrook Road site, 1990s Black Country development area, UK

Several Satisfactory redevelopment of a brownfield site Site formation and completed incorporated with an open-pit recovery of unworked building successfully buildings. 26 ha coal. Engineered fill placed to a depth of 35 m (Hake, completed 2007)

4

Telford Forge, UK

Retail buildings totalling 30 000 m2

Satisfactory redevelopment of a brownfield site. Successfully Following extensive SI, existing fill was re-engineered completed to meet current standards (Hake, 2007)

5

Donnington Wood, 2005 Telford, UK

Storage packing and distribution, including high bay storage

A coal waste site had been reworked and treated Successfully (modified dynamic compaction) by a development completed authority. Following comprehensive SI, decision taken that current settlement tolerances required reengineering by vibro-columns and high-energy dynamic compaction

6

China

2006

20 000 m2 workshop for precisionengineered products

Non-engineered fill placed over existing ponds filled with untreated, soft organic silty clay

7

China

2009

30 000 m2 depot Previous non-engineered fill placed over alluvial flats for UK-based plc by development authority. Fill re-engineered and alluvial soils strengthened by cement-soil piles to reduce settlement to tolerable standards

8

China

2009

20 000 m2 depot Building platform designed to have a graded gravel Material rejected and layer (75 mm–0). Material delivered to site was up to removed from site 400 mm in largest dimension, and was unprocessed blasted quarry material, despite having a complying certificate as gravel

9

Central Europe

2008/09 Chemical factory with precision machinery and high bay warehouse

1990s

Floor settlement exceeded 350 mm. Currently being jacked back to level using expanding resins Successfully completed

Successfully (a) 6 m high embankment to be constructed over a completed pre-existing valley in loess-derived soils. Engineered fill designed to accept floor loading; column loads to (Figure 17) be supported by vibro gravel columns. Platform design included underdrainage, careful selection and use of materials and benching. Filling carried out by a separate agency in contract to development authority, but supervised by building contractor by agreement. Underdrainage and controlled engineered fill emplaced and quality verified (b) Neighbouring site had non-engineered filled Localised internal platform without positive drainage and rodent scouring failure of activity (ground squirrels) (Figure 18) neighbouring fill slope, later repaired ( continued) 97



No. Location

Date


Outcome

10

G Park, Rugeley, UK

2009

UK-based plc. Storage and distribution, 60 000 m2

Situated on 30 m deep reworked brownfield site. Following comprehensive SI, fill was re-engineered and treated using DC techniques to meet current settlement requirements

11

Scandinavia

2009

22 000 m2 distribution depot

Filled platform constructed in good faith by developer Voids apparent in up to 13 m deep formed in unprocessed rock fill, upper surface of and in layers 2–3 m thick. Basal surface formed by platform plunging large boulders (1 m3 +) into soft silty clays. All said to be compliant with Scandinavian practice

Re-engineering was successfully completed (Figures 19 and 20)

Figure 17. Successful project in Hungary built on 6m engineered fill platform

Figure 19. Re-engineering existing platform using DC

Figure 18. Non-engineered fill with rodent activity

98

Figure 20. G Park, Rugeley, built on re-engineered fill platform



Acknowledgements

Aspects, 2nd edn. Building Research Establishment, Garston, BRE Report BR424. Clayton CI (2001) Managing Geotechnical Risk: Improving productivity in Building and Construction. Prepared for ICE under the DETR Partners in Technology Programme. Thomas Telford, London. Godfrey PS (1995) Control of Risk: A Guide to the Systematic Management of Risk from Construction. Construction Industry Research and Information Association, London, CIRIA Funders’ Report FR/CP/32 (republished as CIRIA Report No. 125, 1996). Government of Hong Kong SAR (2004) Fill Slope Recompaction: Investigation, Design and Construction Considerations. Civil Engineering and Development Department, HKSAR, GEO Technical Guidance Note No. 7 (TGN7). Hake SS (2007) Applications in engineering geology. Engineering Geology and Geotechnics at Portsmouth: Alumni Conference Proceedings, pp. 16–19. Hong Kong Government (1977) Report on the Slope Failures at Sau Mau Ping, August 1976. Government of Hong Kong, pp. 6–53 (Reprinted 1995 and 2002 as GEO Report No. 86, Geotechnical Engineering Office, Hong Kong). Knill JL (1982) Moderators’ report on engineering geology and rock mechanics. Proceedings of the 7th South East Asian Geotechnical Conference, Hong Kong 2, 161–162. Trenter NA and Charles JA (1996) A model specification for engineered fill for building purposes. Proceedings of the Institution of Civil Engineers – Geotechnical Engineering 119(4): 219–230.

The author should like to thank Raymond Bates, Retired Construction Director Government of HKSAR, for recollections of the Sau Mau Ping disaster; Gazeley UK Ltd, for the use of photographs of successfully completed projects; GSE Group SA Avignon and UK GSE Ltd for access to information on successfully completed projects, and the provision of one photograph; Simon Hake, Director of Wardell-Armstrong LLP, for information on successfully completed projects; Nicolette LeRoux, for prepared figures; Alastair Lumsden, Editorial Manager of the Journal of Petrology, and formerly Principal Lecturer at the University of Leeds, for confirmation of personal communication, and other advice on publication; Keith Nicholls, Principal Engineer with Geotechnics Ltd, for constructive comments on the recommended procedure; Dr Graham Smith, Director, Geosynthesis Ltd, for useful comments and suggestions, and two photographs used in the publication; and Ir. William S. K. Wong, Chief Geotechnical Engineer, Government of HKSAR, who provided up-to-date references from Hong Kong. REFERENCES

Burland JB (1987) Nash Lecture. The teaching of soil mechanics:

a personal view. Proceedings of the 9th European Conference on Soil Mechanics and Foundation Engineering, Dublin 3, 1427–1447. Charles JA (1993) Building on Fill: Geotechnical Aspects. Building Research Establishment, Garston, BRE Report BR230. Charles JA and Watts KS (2001) Building on Fill: Geotechnical

WHAT DO YOU THINK?

To discuss this paper, please email up to 500 words to the editor at [email protected]. Your contribution will be forwarded to the author(s) for a reply and, if considered appropriate by the editorial panel, will be published as a discussion in a future issue of the journal. Proceedings journals rely entirely on contributions sent in by civil engineering professionals, academics and students. Papers should be 2000–5000 words long (briefing papers should be 1000–2000 words long), with adequate illustrations and references. You can submit your paper online via www.icevirtuallibrary.com/content/journals, where you will also find detailed author guidelines. 99

The design and construction of filled building platforms (engineered ﬁll constructed over low or uneven terrain)

Recommend Documents