PAPER

"~eouI: ine( esi~ n e 'earl: >

re':aininc wa sDr John Oliphant, Department of civil 8 offshore

engineering, Heriot-Watt University, Edinburgh

Obtain site and project information and define design criteria

)F

Carry out desk and site surveys

IF

Conceptual designCompile catalogue of alternative wall types.

Use tables 1a and Ib for guidelines

Carry out preliminary ground invesbgation

le outline design of earth retaining walls involves the choice of

wall, while the detailed design concerns the numerical calcula-tions necessary to ensure the safety of the chosen wall. The out-

line design is less understood and documented by geotechnical engi-neers than detailed design. However, wall selection forms a crucialpart of the overall design process and hence should be given muchgreater attention. This paper focuses on the outline design by propos-ing a practical design procedure and targets the factors which domi-nate the choice of wall through the examination of a number of casestudies. The suggested practical design procedure should lead toeconomies of selection as a more informed comparison of alternativeretaining wall types can be made.

Introdnc5onEarth retaining walls are commonly used to support soils andstructures to maintain a difference in elevation of the ground surface.This difference in elevation can be produced through excavation belowexisting ground level or by earthworking above existing ground level.There is a large variety of retaining walls to choose from and a numberof factors govern the selection. Retaining walls will normally begrouped into gravity walls or embedded walls. Gravity walls are themost common when construction is above ground level and provideretention by their weight alone which can consist of concrete, masonryor selected stone, or by a combination of this material and the soilresting on the heel of the walL Embedded walls, on the other hand, relyon the resistance of the soil in front of the wall of stability. The stabilitycan be enhanced through the use of external support systems such asprops.

The overall design of retaining walls will involve outline and detaileddesign stages. The outline design is primarily concerned with the choicebetween types of retaining wall, while the detailed design concerns thenumerical calculations carried out to ensure the safety of the selectedwall.

The selection of an appropriate retaining wall is particularly difficultif the quality of some of the crucial information is doubtful. Choicebetween wall types will be affected by the soil and groundwaterconditions, the site conditions, construction requirements,environmental factors, the nature and size of project, and will often haveconsiderable cost implications. The prioritising of these factors willreflect the nature of the work, the defined design criteria, and how thedesigner or contractor approaches the design. For an informed choice ofwall, the engineer would need to exercise considerable judgement andexpertise.

The outline design of retaining walls is less understood anddocumented by geotechnical engineers than detailed design. However,the choice of wall is a crucial part of the overall design process andshould be given greater emphasis. This paper concentrates on theoutline design by proposing a design procedure and targeting the factorswhich dominate the choice of wall types through the examination of anumber of case studies. Advice is given on the application of the maintypes of retaining wall and their preliminary stability analysis. A briefdiscussion is also included on work currently taking place to improvethe approach to wall design through the development of a knowledge-based system.

General design philosophyThe overall design of earth retaining walls is a somewhat complexprocess, normally incorporating some degree of analysis but moreimportantly involving the choice of the wall. Knowledge will berequired of the many different types of wall available and this will formthe conceptual design stage. The number and diversity of potential

Preliminary designStep 1

Reduce catalogue to a few alternatives based on selection criteria andpreliminary assessment of stability.

Use table 2 for guidance

Preliminary designStep 2

Carry out comparative cost study and environmental assessment

SELECT RETAINING WALL

RINna1: ~.~ fnnsafftnn far fhaanfffna ffssfgn sf aarfft nnfnfnfng naNL

solutions will be related to the defined design criteria. Recentdevelopments in wall construction techniques such as reinforced soil,bored piles and diaphragm walls, and the use of new materials such asgeotextiles, should form an important part of the catalogue of options.Of the large number of alternatives postulated at this initial stage, manywill be eliminated, perhaps without explicitly being considered whenevaluated against the defined design criteria, The breadth of this rangeof alternatives precludes the use of detailed and complex analyses butwould include knowledge of the project and site as obtained from thedesk study and site survey. Any sort of ground investigationinformation at this stage would rarely be required. In the earlyconceptual phase of design, the engineer would certainly benefit frombeing able to recall examples of existing successful designs and listingpoints that should be considered at a later stage.

Preliminary design follows the conceptual design stage and wouldnormally be carried out in two steps. Step one involves eliminating theleast feasible wall types from the catalogue of wall types producedearlier. This selection process will not be straightforward; it will requireconsiderable judgement and expertise on the part of the engineer. Inevaluating the alternatives, the engineer's decision will be influenced byfactors such as:

~ ground, groundwater and tidal conditions~ proposed height and ground topography~ availability of materials (eg backfill) and specialist equipment~ construction space available~ ground movements (susceptibility of adjacent structures) andexternal loads~ design life and maintenance requirements~ underground obstructions~ appearance~ confidence in design and construction.

The final decision will not normally be influenced by a single factor butbe based more on a number of the factors given above. It would beappropriate at this step to assess the stability of the wall types f'rompreliminary calculations covering the main limit states.

At the end of step one a list of feasible wall types would be producedfor final consideration at the second step. The second step narrows the

GROUND ENGINEERING SEPTEMBER 1997 53

PAPER

Gabions

nnI ~ 'I48~ ssll:— .=:8r~„wana

GRAVITY MASS CONSTRUCTION Gravity walls are generally convenient for construction from ground level up.WALLS

Mass walls are a simple form of construction. Common for walls up to 2/8m high.Rare above 6/7m. Consider sloping front face for appearance and to aid stability.Large quantity of concrete can be relatively expensive.

Concrete with masonry facing For masonry facing consider differential movement between facing and concreteUnreinforced masonry (brickwork, stone, and ensure facing is adequately tied.

other materials) Gabions have been widely applied to river and tidal works. In large walls gabionscan be tied back to augment stability. Cost effective if suitable stone is readily avail-able.

For low crib walls (<2m) plumb face is suitable while for high walls (>4m) battercan be increased. Used for support to embankments and its rustic appearancemakes it more acceptable in rural situations. Construction requires careful super-vision; can be unsatisfactory if setting out and element tolerances are not carefullycontmlled. The backfill needs to be drained and the infill needs careful selectionsince some granular materials cow out.

SEMI MASS CONSTRUCTION

ConcreteConcrete with masonry facing

Semi mass construction is a compromise between simplicity of mass concrete andlow material content of reinforced concrete. Can be cost effective if reinforcementdetails are kept simple.

REINFORCED CON)TQUCTIONConcrete cantilever ( &~)Counterfort/buttressedPrecast

Masonry (Grouted-cavity; Quetta bond;Pocket type; Hollow blockwork)

Prestressed (Post tensioned diaphragm)

Soil (rigid or flexible facing; extensible orinextensible reinforcing)

Reinforced concrete is the most common form of construction for heights up toabout 8m. For greater heights a counterfeit wall is more appropriate. Compaction offill between counterfeits is difficult and not always satisfactory. Buttressed wallsare seldom used. Cantilever wall is often cheaper than mass concrete but relativemerits are balanced. Complicated reinforcement makes construction slower thansemi mass. Wall geometry reflects stability criteria, availability of space andappearance. Base and stem are better designed with more than the theoretical eco-nomic thickness. Stiffness is increased and steel fixing and concreting are easier.Installation and fixing of pre cast units should be considered. For basement con-struction (excavation below ground level) temporary support is required.

For guidance on brickwork walls see BS5628:Part 1:1992 and Part 2: 1995.Reinforced masonry is suitable for walls over 1.5m, while pre stressed masonry isusually economical for walls over 4m. Both methods provide walls with highappearance qualities and weathering capability.

Reinforced soil structures tend to be Qexible and have a large tolerance for move-ment, ideal for sites with poor ground near the surface. Construction times are gen-eraffy shorter than for conventional structures with plant and material require-ments being minimal and therefore well suites to sites with limited access. Metallicreinforcing strips can be replaced with geotextiles, creep and durability a concernfor permanent walls. Reinforced soil structures less likely when strips interferewith boundaries and obstructions. Full guidance is given in BS&006:1995.

list down further to a preferred solution based on environmentalconsiderations and on a relative cost comparison. During thepreliminary design stage the engineer would tend to rely on informationobtained primarily at the preliminary ground investigation stage.

The conceptual and preliminary design stages represent the outlinedesign of the retaining wall which forms a critical part of the overalldesign process. The flow of information between these two stages is twoway, reflecting the iterative nature of design where checks andrefinements will take place.

The retaining wall selection process (preliminary design) is lessunderstood and documented by geotechnical engineers than thedetailed design of which analysis forms an important part. For example,a review of BS8002 (Anon') shows that the bulk of the content isconcerned with detailed procedures of design and the control of safetythrough design calculations, but it pays little attention to the selectionprocess ('/, page in a document over 100 pages long). However, it doesoutline the main limit states associated with common wall types andtheir potential application. Similarly, EC7 (Anon') is confined entirely toaspects relating to design calculations in minimising the occurrence oflimit states.

The final stage in the design procedure mainly concerns thenumerical calculations carried out to ensure the geotechnical and, whennecessary, structural safety of the selected wall. Tools of increasingtechnical sophistication can be employed for this. The informationrequired for a full design would be obtained from the main or detailedground investigation.

Online design: Proposed design procedureBased on the general design philosophy outlined above, a procedure forthe outline design of earth retaining waUs is proposed in Figure 1.

For the purposes of identifying the particular requirements of theconceptual and preliminary design stages, it will often be convenient tocarry out the desk and site surveys before the conceptual design stage,

and the preliminary ground investigation stage after the conceptualdesign but before the preliminary design. The continuity of theseinformation gathering stages will depend on the nature, size and type ofproject. For example, in projects where the retaining waU forms part ofa larger project, then the stages will often be carried out together, andin relatively small projects the preliminary ground investigation maynot be required. In most cases, however, the two stages would beconsecutive.

The use of Tables 1(a) and 1(b) provides a useful reference (initialcatalogue) during the conceptual design stage. Table 1 classifies themain types of earth retaining wall in terms of gravity walls whereconstruction tends to be above ground level, and embedded walls whereconstruction tends to be below ground level. The wide range of walltypes makes selection difficult and hence guidance is given on theirmain application or usage.

The table has been compiled on the bases of the experience of theauthor and on the information obtained from the following sources:BS8002 (Anon '), Hambly', and Fleming et al'.

The catalogue of alternative retaining wall types established duringconceptual design will be reviewed and then refined on the basis of theselection criteria and a preliminary assessment of stability (STEP 1)and the final choice of wall will be based on environmentalconsiderations and a relative cost study (STEP 2). The selection criteriarelate to the factors mentioned previously which affect the choice ofretaining wall.

The choice will not always be obvious nor straightforward, but willdepend on a number of factors. The engineer in this case would rely onexperiences on other retaining wall projects to evaluate the othersolutions to determine the most appropriate. In the next section, fourcase studies are presented which highlight the importance of this earlyevaluation, the main factors affecting the final choice of retaining walland the selection process.

The preliminary assessment of wall stability will involve simple

GROUND ENGINEERING SEPTEMBER 1997

PAPER

%hhfbt~anC " 'of

Wa,'.",".;::,,--,,"."-,'-,:- -,—,~a'aa'-.; '.:,EMBEDDED SHEET PILE CONSTRUCTIONWALLS Cantilever or anchored

TimberSteel

partly driven Remfo ced concrntePrestressed concrete

Propped (temporary)Strutted excavations

excavated

Embedded walls are generally convenient for construction from ground level down,when their high cost is compensated for by the speed of construction and lack oftemporary work. They usually require some form of facing after excavation. Theirselection depends very much on whether or not they are required to retain water, itis expensive to choose a water tight type of wall when it is not required.

Cantilever walls are suitable for moderate retained heights. For flexible steelsheet piles, up to 5m is acceptable, while for stiffer reinforced concrete sections, upto 12m is acceptable. However, deflections should always be checked especially atthe head of the wall. Cantilever not suitable if services or foundation occurs in theactive zone. To reduce penetration length, defiections, and bending moments one ormore levels of anchors or props can be installed. For information on groundanchorage's see BS8081:1989.The choice of using ties or props is complicated, butthey are essential where hard stratum occurs at such a depth that the penetrationof a cantilevered wall is either impracticable or excessively expensive.

SOLIDER/KING PILE CONSTRUCTIONCantilever or anchoredSteelReinforced concrete

(insitu or precast sheeting)

Steel sheet piling has generally been associated with temporary works and used toconstruct quay walls and to support and protect river banks. However, it is nowused to advantage for permanent works with protection/additional thickness pro-vided against corrosion, and has been used successfully on bridge abutment, pump-ing station, basement, underground car park and motorway widening projects.Economy is evident when sheet piles are needed in any case for a temporary coffer-dam. For further information on the design and construction of steel sheet coffer-dams see CIRIA SP95. Finally, it is important not to use too small a section whichcannot be driven into the ground.

Soldier/king piles have been used to support deep, narrow, shallow or wide exca-vations in clays and sands. Excavation in water bearing ground may require specialattention. Method is unsuitable for the exclusion of water.

BORED PILE CONSTRUCTIONCantilever or anchoredContiguousSecant

Contiguous bored piles are often the more economic for small or isolated works.Best in cohesive soils, but may be unsuitable when water level is high on retainedside of wall. Piles can be 'secant'or better water sealing providing the tolerancesof positioning and vertical alignment are sufficient to prevent any gaps.

DIAPHRAGM WALL CONSTRUCTIONReinforced concretePrestressedPrecast

Diaphragm walling may be economic for large or repetitive works. They can be con-structed using heavy rope operated grabs to depths of about 50m, though the diffi-culties of extracting very deep stop ends are considerable.

SOIL NAILING CONSTRUCTION Soil nailing is a relatively recent construction method of reinforcing the ground insitu. The method should be given serious consideration in the permanent supportof excavations. Guidance is given in BS8006:1995.

50

H H

0.25B

I 0.1H

0.4H to 0.6H B = 0.5H to 0.8H

Rgmo 2a:Common propmffons for mass oomanlo gravity walla

GROUND ENGINEERING SEFTEMBER 1997

Rgmn 2b:Commoll Iaa aaIaaaa aaons for maaafaaa aaaaaa ~cia aaaaa oanfgosor waIL

PAPER

Mos — — — ~~ra Hoe~Us .—x:—

B (m)s Surcharge had = 10kWm~

i

Gravityretalnmgwalls

Overturning. Translation; Bearingpressures; Rotation of soil mass; Walldeformations and ground movements.

Embeddedretainingwalls

CantileversForward rotation; rotation of soil mass;wall deformations and ground movements.

Single anchors or tiedRotation about anchor or tie;Rotation of soil mass.

Multi anchored or tiedOverall equilibrium (postulated failuremechanism); Base instability; Walldeformations and ground movements.

H (m)

H H W (m)

Foundation material

4 = 28'

= 1.5H to 2.0Hd = 1.5H to 2.0H

IaeaP. Fleare4:I)ogesII 2:==:,tres)ye chartrera mass Srlrltr waa

Seelg IEFI!Rgaesa Wlhl~~~ror m:—=:-=--Inc~~ wag(Ilxalcaret ssppiss s ~emmet')ISO% IIQIR Rgaia 3IL'hleat'==:: «7 rare:::::=:::gsi nsrsisss ~ss~ wasglee eath eayyett~~).

Uncompacled granular backfill P = 28'o 44

28. j 30.f 32 ! 34 ! 36'! 38' 40 [

H Base wicnI W

1.0 1.40 1.35 1.30 1.25 1.20 1.15 1.101.5 1.90 1.85 1.80 1.75 1.70 1.65 1.602.0 2.80 2.70 2.60 2.50 2.40 2.30 2.2025 3.50 3.40 3.30 3.10 3.05 2.95 2.803.0 4.30 4.10 3.95 3.80 3.65 3.50 3.353S 5.00 4.80 4.65 4.45 4.25 4.10 3.904.0 5.55 5.40 5.20 5.00 4.75 4.55 4.354.5 6.30 6.10 5.85 5.85 5.40 5.20 4.955.0 6.70 6.45 6.20 6.20 5.75 5.55 5.30

42! 44

1.05 1.001.55 1.502.10 2.002.65 2.503.20 3.003.75 3.504.15 4.004.70 4.505.15 5.00

methods of analysis such as limit equilibrium techniques, stress fieldmethods or limit analysis approaches (Potts'). These methods will beappropriate in obtaining first estimates of stability in relation to thegeotechnical limit states given in Table 2.The choice of whether or not toassume undrained or drained behaviour would have to be considered.Preliminary calculations can be shortened by a prudent choice of theinitial wall geometry, where simple changes in shape can improvestability as easily as an increase in weight. Common proportions forvarious types of gravity wall are given in Figure 2, while Figure 3 showsinitial dimensions for the stability evaluation of cantilever and singlepropped embedded walls.

BS8006 (Anon') provides initial dimensions for reinforced soil wallsand abutments(Table 19and Figures 19and 20).

The use of preliminary design charts can simplify the analysis.Oliphant and Corney'ave produced a series of preliminary designcharts for mass gravity retaining walls. A typical chart is shown inFigure 4 where a minimum base width can be established to satisfy boththe overturning and sliding limit states. Other charts have beenproduced to cover other loading conditions and a range of backfill andfoundation material types. The charts are simple to use and provide arapid means of establishing the main wall dimensions for thesubsequentcoststudy.

British Steel'as produced preliminary design tables for steel sheetpiles in cohesionless and cohesive soils. For cohesionless soils,recommendations are given on pile sections and lengths for variousretained heights of both cantilever and anchored walls. For cohesivesoils assuming undrained conditions, recommendations are also givenon pile sections and lengths for various retained heights, but forcantilevers only.

Bica and Clayton'ave produced preliminary design charts forcantilever embedded walls in granular soils based directly onexperimental data. These charts can be used to evaluate the depth ofembedment and maximum bending moment for selected values of planestrain angle of shearing resistance and factor of safety.

Factors amoclng selection: Case Stndy1Howard et al" examined a number of types of retaining wall for themarine esplanade in Liverpool. A wall was required to form apermanent barrier between the River Mersey and the InternationalGarden Festival Site. The main design criteria included a two yearcontract limit, a budget limit of KI.5m, a 100year design life, and a 500msection of wall at a site with difficult tidal conditions.

A feasibility study was carried out on a number of retaining walltypes, each evaluated against the design criteria, and the results areshown in Table 3. It was only through a detailed cost study that the posttensioned precast concrete wall was found to be too expensive. Theauthors stressed the importance of considering the ease of constructionof each wall option, and not being preconditioned by tradition in alwaysopting for the reinforced concrete cantilever retaining wall.

Yonan" has reported on the widening of the M3 motorway from theexisting three lanes of carriageway to four lanes. A number of retainingwalls were considered as the most effective solution for the motorwaywidening. The chosen wall had to satisfy two main design criteria. Firstit had to be environmentally acceptable in terms of landscaping andnoise pollution, and secondly a fast and effective construction methodwas needed to minimise traffi congestion and disturbance.

The preliminary design considered the economies of constructionand environmental impact of gabion, reinforced concrete, sheet pile andreinforced soil retaining walls. Table 4 shows the relative costcomparison for the wall options based on a particular set of parameters.Costings were done for walls constructed to a batter of one horizontal totwo vertical and a retained height of 2m. The reinforced soil wall waspriced for both geogrids and polymer strips as the reinforcement. Alsoincluded in Table 4 is the cost effectiveness ratio (CER) for the reinforcedsoil wall with polymer strips over the wall options. The CER is defined asthe ratio of the cost of the reinforced soil wall with polymer strips to thecost of the alternatives and, hence, the greater the CER the greater the

GROUND ENGINEERING SEPTEMBER 1997

PAPER

era= Ne&leeenroe~ —=——— ~S:: 4ONea Isea—

Steel sheetpfling

In-situconcretereinforced

Pre-castreinforcedconcrete

Post tensionedpre-castconcrete

Massconcrete

Not considered feasible.Difficult to drive sheet piling and guarantee durabilityof steel for 100years in such a hostile environment.Difficult to achieve successfIQ toe anchorage insandstone of varying degree of weathering. Costly totie back head of pile to reduce bending moments.

Considered feasible but impractical.Costly and time consuming to provide watertightcaissons and a delay in reclaiming and landscapingthe garden site. Concerned with the corrosion of thereinforcement in the marine environment.

Considered feasible but impractical.Major problems with levelling, lining and fixing thepre-cast sections in position.

Considered feasible but impracticalFeasible to construct andprovide a durable andattractive solution. Found to be too expensive.

Considered feasible and practical.Found tobe the most cost effective, durable andvisually attractive.

Construction time

Capability ofcontractorMaterials

Wall geometry

Ground and groundwater conditions

Site accessibilityMovements duringconstruction

Environmentallyacceptable duringconstruction

Costs

Preliminarystability analysis

Environmentallyacceptable inintended location

%8@4:::ONInr.: Case Study 1 demonstrated that a simple masssavlilg.The retaining wall with reinforced soil with

polymer strips was finally selected as it imposedminimum environmental impact by restoring thenatural appearance of the motorway slopes andallowing vegetation growth on the wall, and had theadded benefit of reducing noise pollution.

design process followed that of Figure 1 where theGabion 400 1:1.0 conceptual and preliminary design stages wereReinforced concrete 600 1:1.5 performed with a detailed cost study carried out on

1200 1:3.0 two of the wall options.Case Study 2 placed emphasis on the cost of

500 1:1.25 construction and environmental assessment,400 — normally considered later in the outline design

Case Study 3 Sheet pileThe design and constructionof the deep basementfor Reinforced soil:the new headquarters of the United Overseas Bank in GeogridSingapore has been described by Wallace et al". The Polymer strips

process. A clear advantage of using the relativelyinnovative reinforced soil technique was foundcompared to the more conventional retaining walls.The cost effectiveness ratio for the chosen wall is 1:1.5

compared with the reinforced concrete cantilever wall, 1:3.0comparedwith the sheet pile wall, and 1:1.25compared with the reinforced soil wallwith geogrids.

The evaluation of the retaining walls in Case Study 3 was carried outat the preliminary design stage without resorting to any cost orenvironmental study. The factors affecting selection related more to thetwo main design criteria as well as to the lack of contractors withexperience in a specialist construction technique.

The decision to use soil nailing in the fourth case study was basedentirely on the nature of the site and type of project and no explicitconsideration was given to alternatives at the preliminary design stage.

The case studies have shown that the extent of the outline designvaries considerably and appears to be related to the type of project andthe importance placed on environmental considerations. Clearly, whenreferring to Figure 1, the engineer has the option to omit any of thedesign stages but this would only be done with a good knowledge of thesite and the proposed project.

General discussion I condIndonsThe suggested procedure for the outline design of earth retaining walls(Figure 1)should lead to improved understanding of the variety of walltypes available, the range of factors affecting the choice of wall, themethods available for preliminary design and the structured approachtothedesignprocess.

This should result in a more informed comparison of alternatives (egreinforced soil wall vs reinforced concrete cantilever wall) in order tomake economies. The development of a computer-based decisionsupport system encapsulating the above features of the design processwould greatly benefit the engineer. This is the current focus of researchby the author in developing a knowledge based system for the design ofearth retaining walls. There are two main objectives to this researchwork. First, to develop methodologies for eliciting and analysing expertopinion on the wall selection process. This expertise will comeprimarily from a wide range of contractors in the UK. Secondly, toprovide amore rigorous and structured approach to the outline design of

site is underlain by up to 30m of soft marine clay andthe water level is near ground surface. Theexcavation was typically 10m to 12m deep but reached16m in places. The chosen permanent support to thesides of the excavation would have to act as a deep groundwater cut offand minimise adjacent ground improvements. Diaphragm walls werechosen as they satisfied these two criteria and there were severalcontractors in Singapore who had gained considerable experience ofdiaphragm walling during construction of the Singapore Mass RapidTransport System (Hulme et al").

Steel sheet piles were considered unsuitable due to the large groundmovements associated with their use, although no attempt was made toestimate these movements at the preliminary design stage. Contiguousbored piles were unfeasible as they were not sufficiently watertight.Secant bored piles were not considered beyond the conceptual designstage as there were no contractors in Singapore who specialised in thistechnique.

Case Study 4This case study involved the steepening of a natural slope in a ruralsetting during realignment of the A7 Edinburgh-Carlisle trunk road atColterscleuch, south of Hawick. An ancient monument was located atthe top of the natural slope. Two conditions were placed on the chosenretention system. First, the ancient monument was located close to thecrest of the slope and had to be protected throughout construction.Secondly, the steepened slope had to blend in with the naturalsurroundings. Soil nailing was the only method considered to satisfythese conditions and the one solution to be proposed at the preliminarydesign stage. The soil nailing construction afforded minimal grounddisturbance and when a geofabric was used to face the slope, thissupported the growth of grass and provided an acceptableenvironmental solution.

Discussion of case studiosThe case studies have shown that the selection of earth retaining walls isinfluenced by many factors involving design, construction,environmental and economical considerations. Table 5 summaries thefactors considered explicitly at the preliminary design stage whichinfluenced the final selection for each case study.

+ Cost effectiveness ratio

concrete wall can be cost effective, durable andvisually attractive. It also shows that simplicity in

InnogSle'::- ..4lig- ~.. construction especially under difficult conditions isgANnat 'enerally considered to be cost effective. The outline

GROUND ENGINEERING SEPTEMBER 1997

PAPER/MEETlNI REPORT

retaining walls.The primary role of the support system will be to aid the experienced

engineer in matters relating to the outline design of retaining walls. Thesecondary role of the system will be in training graduates and lessexperienced engineers. Research work is being funded by the EPSRCBuilt Environment.

a~INiKlmawg ~w Smmaaw

The author is grateful to A Clark of the Scottish Office DevelopmentDepartment, National Roads Directorate, and to P Frankland of BordersRegional Council for providing information on the A7 Colterscleuchimprovement and for giving permission to publish it.

twssmmsmmsa1 Anon, BS800Z Code of Practice for Earth retaining structures, BSI Lmfoni Wood

Milton Keynes, England, 19942.Anon, Eurocode 7 Part I:~cal Design, General Rules, 4th Version, February,

19933.Hambly, EC, Bridge foundations and substructures, BREReport, HMSO, 19794.Fleming, WGK, Weltman, AJ, Randolph, MF &Elson, WK, Piling Engineering, Surrey

University Press, Glasgow, 19865. Potts, DM, The analysis of earth retaining structures, Retaining Structures, pp 167-

185,Thomas Telford, London, 19936.Anon, BS8006:Code of Practice for Strengthened/reinforced soil and other fills, BSI,

Linford Wood, Milton Keynes, England, 19957. Oliphant, J & Corney, N, The design of gravity retaining walls, Internal Report,

Department of Civil &Offshore Engineering, Heriot-Watt University, Edinburgh, 19928.British Steel General Steels, Piling Handbook, sixth edition, Scunthorpe, 19889.Bica, AVD & Clayton, CRI, The preliminary design of free embedded cantilever walls

in granular soil, Retaining Structures, pp 731-740,Thomas Telford, London, 199310.Howard, J, Curtin, WG &Pope, LS,The Structural Engineer, V67, NI, January, pp 3-9,

198911.Yonan, SJ,Geotextile retaining walls for motorway widening, Retaining Structures,

pp541-548, Thomas Telford, London, 199312.Wallace, JC, Ho, CE & Long, MM, Retaining wall behaviour for a deep basement in

Singapore marine clay, Retaining Structures, pp195-203, Thomas Telford, London, 199313. Hulme, TW, Potter, LAC & Shirlaw, JN, Singapore mass rapid transit system:

construction, Proc. ICE, Pt I, 1989,86, Aug, 709-770.

i,'I )eN all(c ose( caseReportofa meeting entitled "Services under streets: adebate on trenches versustrenchless technology" held by

East Midlands Geotechnical Group at Loughborough

University on19 May, by Richard Bennettof Nottingham

University.

ecting chairman Dr David Chapman of Nottingham Universitystarted the debate by running through the perceived advantagesand disadvantages of trenchless technology. He suggested that

trenchless techniques offered reduced surface disruption and poten-tially lower overall costs than conventional open cut service installa-tion. Disadvantages of a trenchless approach were a higher degree ofrisk and a need for specialist equipment, which created high initialdirect costs.

Alwyn Morgan of Midlands Electricity (MEB) speaking in favour oftrenchless technology, said his organisation installed 1,000km of cablingand services a year, at a cost of R16M to RISM. Approximately 50 / of thecables are installed in unmade ground, the remaining 50/ in madeground, mostly under urban footpaths.

MEB, said Morgan, has a declared policy of using trenchlesstechnology wherever socially or economically justified, the long termmission being the concept of "zero dig". The company employs fourdirectional drilling machines and several hundred impact moles.Contractors are used in addition to the direct labour force.

Morgan covered the development of trenchless techniques at MEB,commencing with no dig trials with Grundomat and Pierce Arrow soildisplacement tools. These proved to be very cost effective for cable roadcrossings and service replacements in domestic properties. In 1991MEBpurchased a Ditch Witch 420 Jet-Track. This was used for river, road,railway and canal crossings and was an economic success, paying foritself within three months of operation.

The experiments continued with the use of moles and jet-tracks alongfootpaths where the exact location of utilities were known. These testswere successful but it became apparent that lack of knowledge about theexact location of services was hindering progress. Ground probingradar (GPR) was used in subsequent trials and found to be effective forthe detection of pipes and voids, but did not provide a complete answer.

Morgan stated that buried services present a unique problem. Notonly are they hidden from view but they constitute a three dimensionalmaze of pipes and cables. When considering laying new undergroundservices, particularly in urban sites, it is imperative to determine theexact location of services.

Morgan went on to describe an innovative system of threedimensional mapping currently under development by MEB. Existingtwo dimensional records are first studied to determine the extent ofpossible site congestion, followed by an electromagnetic locator survey

to pinpoint services by plan and depth. EML is used to detect serviceswith a metallic content, and a signal emitting sonde for non-metallicpipes or ducts. The data is collated into a three dimensional databasewhich can be interrogated for cross-sections wherever desired.

Selective cross-sections can also be taken by GPR to find plastic pipes,mass concrete or rock horizons. EML and GPR surveys can be comparedto identify anomalies, and in this way the location technique can beoptimised, both in terms of efficiency on site and overall costs. Testshave compared three dimensional mapping cross-sections with trialpits, with considerable success. Development has progressed to the stagewhere installation crews are now using the system to direct cableinstallation. The ultimate goal of three dimensional mapping would beto allow the system to directly steer the rig.

Morgan emphasised that the true benefits of a three dimensionaldatabase are realised as a planning tool. Service routes can be effectivelyplanned and decisions made on the optimum cable installationtechnique to be employed: no dig, open cut, narrow trenching or acombination of all three.

He went on to describe the potential financial savings associated withtrenchless technology used in conjunction with three dimensionalmapping. The example given was a three year linear rate cableinstallation contract, finalised by MEB in 1995. Figure 1 shows thetypical annual high and low voltage cable installation rates, and thecomparison of open cut to no dig contractor prices. The differencebetween these rates offered the opportunity to introduce threedimensional mapping (and so reduce risk by gaining a better knowledgeof the ground), and still reduce overall costs.

A number of projects have now been undertaken in urban areas of theWest Midlands using the system, and are showing savings of 20% overprojected costs with open cut. Without the benefit of three dimensionalmapping, ground uncertainty in all of these projects would lead to theuse of trenching, with the associated disruption and social costs.

Current developments associated with the three dimensionalmapping system include a research project being undertaken withDavid Chapman at Nottingham University to investigate the forcesexerted on existing services by impact moles and directional drills. Thiswill allow the development of an understanding of the effects of usingno dig techniques in ground congested with services. A possible goal ofthis work is to produce guidelines for recommended safe distancebetween existing and new trenchless services in various types ofgI'on nfl.

Morgan concluded by reiterating the cost savings created bycombining impact moling and directional drilling with a threedimensional mapping technique. He emphasised MEB's commitment todeveloping this approach and specifying it as a standard cableinstallation method for both its direct labour and contract installationcrews.

Putting the case for trenches, John Banyard of Severn Trent Waterargued against trenchless technology on the basis that cost overruns intrenchless projects were higher than for conventional projects.

He started on the premise that engineers were often unsuccessful inestimating the costs of construction projects, particularly thoseinvolving work in the ground.

As an example, Banyard pointed out that the Channel Tunnel, hadcost nearly two and half times the original estimate of 24bn, and isviewed by investors as an unmitigated disaster. Huge cost overrunsoccurred on the project, even with the balance sheet being restructuredon several occasions.

GROUND ENGINEERING SEPTEMBER 1997