220527412 geotechnics for the structural engineer

8/17/2019 220527412 Geotechnics for the Structural Engineer

1/156


2/156

Welcome to the sixth edition of the Atkins Technical Journal which again showcasesthe impressive breadth and depth of the technical solutions we offer our clients. Thisedition is strengthened further by the inclusion of papers for the first time from ourcolleagues in Atkins North America.

Sustainability features heavily throughout this edition, confirming that it is as the heartof all we do whether it be planning, designing or enabling infrastructure solutions for

our clients. Some salient examples include:Planning - our work to produce a BIM-style tool to optimise urban masterplanning forsustainability and decarbonisation; the production of holistic energy plans for the UAE;our advice to the Scottish Government to optimise carbon reductions in its transportpolicy; and our advice on water resource management plans to protect and restore thehydrological system in Winter Haven, Florida to meet increasing population demand.

Designing - our production of steel bridge design charts to optimise design for ourclients and industry as a whole; our design of the architecturally impressive AlmasTower, Dubai, with equally impressive design efficiency; and our publications on staycable vibration and fatigue to prevent the need for future expensive remedials on largebridges around the world, as continues to occur at present.

Enabling - our use of GIS and information management to drive efficiencies andget it “right first time” on projects like Crossrail; our expert advice on translocating

wildlife habitats to mitigate the environmental effects of construction; our creationof new wetland habitats in the London 2012 Olympic Park; and our evaluation anddeployment of Open Road Tolling to reduce driver accident statistics and provide saferhighways.

The above summary merely scratches the surface of what we do and what is containedin Technical Journal 6. I hope you enjoy the selection of technical papers included inthis edition.

Chris HendyNetwork Chair for Bridge Engineering

Atkins


3/156


4/156

Technical JournalPapers 0 - 0

081 - Asset ManagementCorrosion mitigation of chloride-contaminated reinforced

concrete structures: A state-of-the-art review

082 - Asset ManagementStochastic based lifecycle planning tool for ancillary pavement assets

083 - Carbon Critical DesignCarbon Critical Masterplanning

084 - Carbon Critical DesignMitigating transport's climate change impact in Scotland: Assessment of policy options

085 - Energy

UAE holistic plan: Kadra power infrastructure development086 - Energy

Creep-fatigue life assessment of reheater tubes at detuning strap connections

087 - GeotechnicalAssessment of a bridge pier pile foundation subjected to bearing replacement

088 - Intelligent Transport SolutionsDeployment and safety benefits of Open Road Tolling for mainline toll plazas in Florida

089 - StructuresThe structural design of Almas Tower, Dubai, UAE

090 - StructuresPreliminary steel concrete composite bridge design charts for Eurocodes

091 - StructuresFatigue analysis of stay cable mono-strands under bending load

092 - SustainabilitySustainable Water Resource Management Plan

093 - Technical Knowledge & Information ManagementGIS and Information Management on Crossrail C122 Bored Tunnels contract

094 - Technical Knowledge & Information ManagementTechnical Networks - The essential framework fordisseminating knowledge from projects to industry

095 - Transport PlanningPedestrian comfort guidance for London

096 - Water & EnvironmentRe-engineering the Mississippi River as a Sediment Delivery System

097 - Water & Environment

Translocating wildlife habitats: A guide for civil engineers098 - Water & Environment

Delivering wetland biodiversity in the London 2012 Olympic Park

5

13

21

25

35

41

49

57

65

75

83

99

111

123

131

137

143

151

Technical Journal 6Papers 081 - 098


5/156


6/1565

Abstract

A S S E T MANA

GE ME NT

81Corrosion mitigation of chloride-contaminated reinforced concretestructures: A state-of-the-art review

Introduction

The problems of concretedeterioration due to corrosion of steelreinforcement and/or pre-stressed/ post tensional systems in concretestructures are worldwide, costingnations billions of pounds, equivalentto around 4 to 6% of Gross DomesticProduct (GDP). A recent cost ofcorrosion study4 estimated the annualcost of corrosion on US highwaybridges to be at $8.3 billion overall,with $4.0 billion of that on the capitalcost and maintenance of reinforcedconcrete highway bridge decksand substructures. In the UK theDepartment of Transport’s estimateof salt-induced corrosion damage is atotal of £616.5 million on motorwayand trunk road bridges in Englandand Wales. These bridges representabout 10% of the total bridgeinventory in the country (Wallbank,1989). There are no simplistic models

to predict the rate of deteriorationdue to corrosion. However, wehave sufficient understanding ofthe corrosion mechanisms andconcrete deterioration processes.

With the development of various NonDestructive Testing (NDT) assessmenttechniques and the recent advancesin protection and rehabilitationmethods, a large percentage ofthese costs could be reduced.

This paper briefly describes thefundamental principles of corrosionof steel reinforcement in concrete.

The paper then gives an overviewof currently available methods ofprotecting corrosion damagedstructures. Finally, the recentadvances in cathodic protectiontechnology for reinforced concretestructures are discussed.

Corrosion mechanism

of steel in concrete

The corrosion of steel reinforcementin concrete is an electrochemicalprocess involving two equal, butopposite, reactions. These areanodic, or oxidation reactions

(eg Fe Fe++ + 2e-)

and cathodic or reduction reactions.(eg O

2 + 2H

2O + 4e- 4OH-)

Concrete has the inherent ability toprotect steel against corrosion. This isdue to the high alkalinity of concrete,ranging between 12.5 and 13.7,imparted by the chemical constituentsof the cement, in particular calciumhydroxide Ca(OH)

2. In this alkaline

environment, a thin film of oxideor hydroxide such as ferric oxide,Fe

2O

3, is formed on the steel surface

rendering the steel ‘passive’ i.e. the

corrosion rate becomes insignificant.

Amey

Principal Engineer

Highways & Transportation

Atkins

Coventry University

Materials Engineer

Reader in CivilEngineering Materials

HomayoonSadeghiPouya

EshmaielGanjian

Sunil C DasCorrosion is a worldwide problem which costs nations billions of pounds.Although corrosion is not a new problem, awareness of it associated with civilengineering structures, particularly reinforced/pre-stressed concrete highwaybridges, multi-storey car-parks and buildings etc. it is relatively new. Corrosionis insidious in nature. Steel corrosion in concrete is only apparent when it isquite advanced. It manifests itself progressively in the form of ‘rust’ stains,cracking, delamination and finally spalling with exposed and corroding steelreinforcement. Proper application of available science and technology cansave a large amount of waste due to corrosion. Over the last two decadesa number of corrosion mitigation techniques have been developed. Someare more successful than others. Cathodic protection is the only proven

technique to stop corrosion of steel in chloride contaminated concrete.


7/1566

A S S E T M

A N A G E M E N T

81

reinforcement corrosion in chloridecontaminated concrete arestochastic in nature. As a result, the

equations to predict life (service lifeor residual life) need to considervariables such as cover thickness,diameter of the reinforcement,chloride content, water-cementratio, temperature, relative humidity,oxygen concentration etc. Furthercomplications arise due to thedifficulties in defining the ‘servicelife’ of the structure or structuralelements (which depend on thetype of structure, element's functionwithin the structure, failure mode

and consequence, and maximumacceptable damage). Recently,two research groups NationalCooperative Highway ResearchProgram (NCHRP)16 and the projectfunded by the European Communityunder the BRITE/EURAM Programme4 have produced ‘User’s Manuals’on predicting the service life/ residual service life of corrosion-damaged concrete structures/ elements. Detailed discussion ofvarious models to predict the residuallife of chloride induced corrosion-

damaged reinforced concretestructures is outside the scope of thispaper. However, understanding ofcorrosion and concrete deteriorationprocesses has led to developmentof various techniques for mitigatingreinforcement corrosion and concreterehabilitation. These are brieflydescribed in the following sections.

However, this protection mechanismmay break down as a result of oneor more changes in the concrete's

chemistry, the most commonand important factors being:

Loss of alkalinity in the concrete(i)

Penetration of aggressive ions(ii)to reinforcement depth

A combination of both factors.(iii)

The main offending ion for thebreakdown of passive film on steelreinforcement is chloride in concrete.The chloride acts as a catalyst foroxidation of iron by taking an activepart in the reaction. According to

Uhlig15

it oxidises the iron to formthe complex ion FeCl3- and draws

this unstable ion into solution, whereit reacts with the available hydroxylions to form Fe(OH)

2. This releases

the Cl– ions back into solutionand consumes hydroxyl ions, asseen in the following reactions:

2Fe + 6Cl- = 2FeCl3- + 4e- (1)

Followed by:

FeCl3- + 2OH- = Fe(OH)

2 + 3Cl- (2)

The electrons released in oxidationreaction flow through the steel to thecathode surface. This process wouldresult in a concentration of chlorideion and a reduction of the pH at thepoints of corrosion initiation, probablyaccounting for the process of pittingcorrosion. The lowered pH at thesesites contributes to the continualbreakdown of the passive oxide film12.

Alternative reactions forcomplex formation are:

Fe2+ + 6Cl- = FeCl6 + 4e- (3)

or

Fe3+ + 6Cl- = FeCl6 + 3e- (4)The above reaction removes ferrous(Fe3+) ions from the cathode area,allowing them to be deposited awayfrom the bar, through the reaction:

FeCl6 + 3e- + 2OH- = Fe(OH)

2 + 6Cl- (5)

This reaction produces rustand releases chloride ion forfurther reaction with ferrousions. In engineering situations,the electrochemical reactions ofthe corrosion process are morecomplex than described above.

Detailed commentary on thecorrosion mechanisms is outsidethe scope of this paper.

Deterioration

processes of reinforced

concrete structuresThere is now sufficiently detailedunderstanding of corrosion inducedconcrete deterioration and it isbecoming important to predict (ifpossible) the residual life of thestructure/structural elements inorder to inform decision making onwhether the structures/elements arelikely to maintain their structuralintegrity or serviceability until theend of a specified ‘design life’. Suchpredictions are also useful in analysing

the cost-effectiveness of the differentrepair or rehabilitation strategies.

A brief literature survey has identifieda number of models for estimatingthe service life or residual life ofconcrete structures/elements withregard to reinforcement corrosion.All these models are constructedaround a specific structure, but allthe models have a common startingpoint i.e. the prediction is based onthe damaged model first proposedby Tuutti18, where the service life of a

corrosion-damaged concrete structureis described as a two-stage process:

An initiation stage, in which(i)corrosion initiates oncesufficient quantities of chloridesreach the reinforcement

A propagation stage, in which(ii)the extent of corrosion damagebuild-up reaches a ‘limit state’ i.e.it is the time when the probabilityof failure becomes unacceptable.

The service life can be expressed,according to Tuutti18, as:

T = ti + t

corr (6)

where:

T = service life

ti = initiation time

tcorr

= propagation (corrosion) time.

Many mathematical models havebeen proposed to estimate the valuesof t

i and t

corr1,8,7,3. All models use an

equation to determine the ti, the

initiation time period, by solving theequations based on Fick’s secondlaw of diffusion; but estimatingthe values of tcorr is not simple.

In reality, the concrete deteriorationprocesses resulting from

Corrosion mitigation of chloride-contaminated reinforcedconcrete structures: A state-of-the-art review


8/1567

A S S E T MANA

GE ME NT

81

Corrosion mitigation and

concrete repair strategy

Highway and building structuresworldwide are deteriorating at anever-increasing rate. The option forrepair of damaged/deterioratingconcrete structures depends onthe nature of the problem(s). Thechoice of rehabilitation and repairtechnique and material would bedetermined from full understandingof the underlying cause(s) of theproblem. With regard to causes;perhaps one major cause of concretedeterioration is corrosion of steel

reinforcement in concrete. Thiscould lead to structural weaknessdue to loss of cross-section of thesteel reinforcement or pre-stressingwire17. These may be groupedunder the following headings:

Traditional concrete replacement•

Electrochemical repair methods•

Corrosion inhibitors.•

All corrosion mitigation methodscan be applied to both structures tobe built exposed to environmentswith risk of chloride contaminationand existing chloride contaminatedstructures. The choice ofappropriate repair solution willdepend on a number of factors.The short, medium, and long-term operational requirementswill dictate to a significant degreethe maintenance strategy.

Traditional concrete replacement

Traditional ‘patch repair’ consistingof removal of defective concretein an attempt to eliminate thecause of the problem is one option.Current practice specifies that thedefective concrete is ‘broken backto a sound alkaline base’, and anysteel reinforcement should be fullyexposed to approximately 25mmbehind the bar over a lengthgreater than its corroding length,and thoroughly cleaned. Two keyphrases in the paragraph aboveshould be noted. These are: ‘brokenback to a sound alkaline base’ and‘thoroughly cleaned’ in reference

to corroding reinforcement. Inpractice, the above two conditionsmay not easily be met. Furthermore,a potential problem with patchrepair using cementitious materialsis to prevent subsequent ingress ofpollutants, including chloride ions.

To overcome this problem the patchrepaired structure can be coated usingvarious proprietary surface coatings

in order to prevent further ingress ofaggressive ions from the environment(Zhang and Mailvaganam, 2006).

Further, in order to prevent or retardthe ingress of atmospheric pollutants,including chloride ions or oxygeninto concrete, the coating systemneeds to be completely ‘pin-hole’free; at the same time the coatingmust be ’breathable’ i.e. must havesufficient permeability to watervapour to avoid water vapour pressurebuild up behind the coating causing

blistering and subsequent failure.In effect, the aim is to control, orat least retard, the rate of furthercorrosion. Along with the controlof corrosion, other strengthening/ rehabilitation methods are used, if sorequired. These methods consist ofproviding additional reinforcement,extra external pre-stress, replacementof damaged structural members etc.The details of such methods are basedon assessment of strength, and areattended by the bridge engineer10.

In summary, traditional patch repairis a short-term remedy which canbe carried out to delaminated andspalled areas. Conventional patchrepair of corroded concrete structuresinevitably introduces 'incipient anode'effect. This is due to the differentelectrochemical behaviour of steelreinforcement in the 'new' concreterepair material and the surrounding'old' but sound concrete (which maystill be contaminated with chloride).The newly patched area (chloride free)becomes the cathode (less negative

potential) and the neighbouring areasbecome the anode (more negativepotential) and start to corrode.Conventional patch repair treats onlythe symptoms not the cause andthe incipient anode effect makesthis repair a never-ending process.

Electrochemical methods of repair

Electrochemical repair methods such

as cathodic protection (CP) or chlorideextraction (CE) may also be used toarrest the corrosion process. Both CPand CE require an active electricalcircuit to be established which forcesthe steel reinforcement cage tobecome cathodic (non-corroding)by providing an external anode(corroding). Cathodic protectioncan be applied in two ways: (a) byimpressed current CP system (ICCP)where CP uses a permanent externalanode connected to an electricalpower supply e.g. transformer-

rectifier and (b) the second approachis termed Sacrificial Anode CP system(SACP) which uses a metal anode(such as zinc) with a higher naturalgalvanic potential than that of steelto establish the necessary drivepotential directly connected to thesteel structure to be protected13.

Chloride extraction is similar toimpressed current CP in that anexternal electricity supply is requiredto drive the process; however, ananode is supplied as a series of

surface mounted panels containing anelectrolyte. The drive voltage for CE isvery much higher than CP as the aimis to draw the negative chloride ionsaway from the reinforcement towardsthe anode and out of the concrete.

Both CP and CE may require someinitial minor concrete replacementrepairs to those areas which aredelaminated, as the current pathwould be inhibited by the crackedconcrete. An additional requirementof the concrete specification for

concrete repairs where CP or CEis to be used is that the resistivityof the concrete must be kept lowsuch that an electrical current canbe passed through the concrete.

An advantage with electrochemicalrepair methods is that those areasof concrete contaminated withchloride do not need to be brokenout. This can result in significantcost savings which more than offsetthe cost of the system installation.

The durability of repairs using

cathodic protection has been wellestablished, provided the systemsare actively monitored6. Advancesin remote monitoring technologyhave reduced monitoring costssignificantly in recent years. The



9/1568

A S S E T M

A N A G E M E N T

81

establish passivity of the steel. Thiscan give rise to an increased riskof alkali silica reaction if reactive

aggregates are present in theconcrete. The higher the impressedcurrent the greater the risk of alkalisilica reaction, therefore, CE inparticular must be chosen with care.

durability of repairs using CE isless well established. The efficiencyof the CE to remove chlorides

from around the reinforcementto establish passivity of thereinforcement can vary significantlyfrom structure to structure. CEmay be required at intervalsduring the life of the structure as

the remaining chlorides migratetowards the steel reinforcement.The advantage of CE over CP is the

short duration of the repair works6.

Impressed current systems alsogenerate hydroxyl ions around thereinforcement, which raise thealkalinity of the concrete surroundingthe reinforcement helping to re-

Figure 2 - Circuit diagram for a Sacrificial Anode CP (SACP) System (with super-imposed polarisation curves)

Figure 1 - Circuit diagram for an Impressed Current CP (ICCP) System (with super-imposed polarisation curves)

CATHODIC REACTION

O2 +2H

2O + 4e- = 4OH-

ANODIC REACTION

2Fe = 2Fe2+ + 4e-

2 Zn = Zn2+ +4e-



10/1569

A S S E T MANA

GE ME NT

81

Furthermore, the benefit of CE isconsidered to be limited to short tomedium-term and would require anumber of repeat applications overthe design life of the structure.

In recent years an alternativeimpressed current repair techniquetermed ‘electro-osmosis’ (although

there is some debate as to whetherthis is an accurate description of theactual process which takes place) hasbeen trialled on chloride contaminatedconcrete (Chess, 1998). The processworks by drawing moisture awayfrom the rebar thereby increasingthe local resistivity and significantlyreducing the rate of corrosion. Theeffectiveness of this technique isvery much an unknown at this time,although limited trials have beenundertaken by the Highways Agency.

For a longer-term solution forcontrolling corrosion, application ofcathodic protection is consideredto be the ‘only technique’ provento stop/mitigate on-going corrosionof steel reinforcement. This isparticularly the case for chloride-

contaminated concrete.

Cathodic protection for reinforcedconcrete has generally been limitedto impressed current systems dueto the relatively high resistivity ofconcrete. More recently sacrificialanode materials have been developedwhich may be applied to reinforced

concrete where particular criteria canbe met. A typical circuit diagrams(schematic) of an impressedcurrent cathodic protection system(ICCP) and a sacrificial cathodicprotection system are shown inFigures 1 and 2 respectively.

The design concept of an impressed-current cathodic protection systemwould consist of the following:

System hardware, consisting of(a)a power source and ground-bedsystems plus associated cabling,

junction boxes, conduit etc. Control and monitoring(b)system, consisting of feedbackcontrol equipment and anarray of embedded referenceelectrodes plus cabling, junction boxes conduit etc.

The most important element ofany successful cathodic protectionsystem is the design of an effectiveanode system to distribute thenecessary protection currenteconomically and efficiently to thereinforcement. Also, it must beeasy to install and possess longterm durability. Other components(e.g. power supply/monitoringequipment etc.) of the CP system canthen be selected to suit the anodesystem, the prevailing corrosionconditions and the environment.

Over the last 30 years there havebeen considerable advances anddevelopments in anode materialsand anode system design withreal possibility of ‘pick n mix’cathodic protection system(s) forabove ground reinforced concretestructures. Anode systems currently

available are given in Table 1.

Table 1 - Impressed current anode types and characteristics13

Anode Type

Long Term

Current

Density per

m2 of anode

mA/m2:

Long Term

Current

Density per

m2 concrete

mA/m2:

Typical

Anode LifeEstimate

(years)

Suitable

for Wet

Structures

Suitable for

Running

Surfaces

Dimensional

& WeightImpact/

Installation

PerformanceQueries

Typical AnodeCost £/m2

Conductiveorganiccoatings

20 20

Max

10-15 No No No

Painted

Someunprovenproducts

20-40

Sprayed zinc 20 20

Max

10-15 Possibly No No

Thermal spray

Consumption

rate

Health & Safety

60-100

Mixed metaloxide (MMO)

coatedtitanium meshand grid in

cementitiousoverlay

110-220 15-110

Varying grades

25-120 Yes Yes Yes

In circa 25mmoverlay

Overlay quality

control

60-100 including

overlay

Discrete Pt/ Ti or MMO Tianodes, withcarbonaceoussurround

800fromcarbonaceoussurround

Circa 10-110subject todistribution

10-20 Yes, not tidal Yes No

Placed in pre-

drilled holes

40-100

Discreteanodes incementitioussurround.MMO Ti or

conductiveceramic

800 Circa 10-110subject todistribution

20-50 Yes Yes No

Placed into

holes or slots

40-100

Cementitious

overlayincorporating

nickel platedcarbon fibrestrands

20 20

Max

10-20 Probably No Yes

Sprayed, circa8mm thick

Limited

experience30-60



11/15610

CP does not restore lost steel, butprovided the steel has sufficientreserves of strength, CP can provide a

cost effective solution. Even when thestrength is inadequate it is possible,in many cases, to combine CP withstrengthening. With a well designedand installed CP system, the costs ofoperation and maintenance wouldbe extremely low. It is now wellrecognised that in most cases cathodicprotection can provide a cost effectivesolution to stop corrosion and theimportance is acknowledged withcodifying by a number of nationaland international standards5.

The performance of the installedcathodic protection systems wouldbe monitored using embeddablereference electrodes and othermonitoring probes. All referenceelectrodes could be integratedinto a Monitoring Unit and couldbe interrogated either manuallyor via an automatic data loggingdevice which could be operatedlocally or remotely. In addition, themonitoring reference electrodes forICCP systems would function tocontrol the system output to provideadequate levels of protection.

Corrosion inhibitors

Corrosion inhibitors work bychemically raising the thresholdof chloride required to initiatede-passivating the reinforcementand initiate corrosion. Corrosioninhibitors have been used for manyyears in the automotive industryand have been demonstrated tobe effective in new build concrete,in particular the use of calciumnitrite as an admixture to freshconcrete. However, the applicationof corrosion inhibitors to existingchloride contaminated concrete hasbeen shown to be less effective2.

Conductive coating anodes includea variety of formulations of carbonpigmented solvent or water dispersed

coatings, and thermal sprayed zinc.Recently, thermal sprayed titaniumhas been used experimentally,with a catalysing agent sprayapplied to the titanium coating9.

Mixed metal oxide coated titaniummesh or grid anode systems arefixed to the surface of the concreteand overlaid with a cementitiousoverlay which can be poured orpumped into shutters or sprayed.

Discrete anodes are usually installedin purpose cut holes or slots in

the concrete. They are either:

Rods of coated titanium in•a carbonaceous backfill

Mixed metal oxide coated tubes•

Strips and ribbon•

Conductive ceramic tubes•in cementitious grout.

More recently, there have beensuccessful experiences with theanode design based on utilising zincrich paint as a sacrificial/impressedcurrent anode material (Das, 1999).

The main characteristic propertiesof the zinc rich coating are:

Coating is easy and safe to(a)use. It is a one-pack compoundcontaining 99.995% purityelectrolytic zinc dust mixedin synthetic resins, pigmentsand aromatic solvents.

Purity of the zinc content is(b)such that there is no lead orcadmium present. The productdoes not contain toluene, xyleneor methyl ethyl ketones (MEKs).Thus the product is non-toxic.

On application it cures to a(c)minimum of 96% zinc contentin the dry film; thus is capable ofproviding full cathodic protection.The coating can be brushed orsprayed on. There is no barrier orinterface between coatings i.e.every coat merges perfectly withprevious coats and therefore canbe topped up time and time againto provide indefinite cathodicprotection at very low cost.

On the steel surface the coverage(d)is approximately 4-5 squaremetres at 30 to 40 microns.

Coating has indefinite shelf life.(e)

Coating can be applied in(f)moist or wet conditions.

Another recent development is the‘Discrete’ Zinc Sacrificial AnodeSystem. This is a proprietary zinc

sacrificial anode unit embeddedwithin a specifically formulatedcementitious mortar and is currentlyavailable commercially. The mainapplication of this anode systemis for localised protection of steelreinforcement within chloridecontaminated concrete by maintaininggalvanic protection in areas adjacentto ‘conventional patch repaired’ areasand thereby prevents the formation ofincipient anodes in neighbouring areasfollowing anti-corrosion treatment

and concrete repair to damaged areas.This anode system is discretelyplaced within the patch repairs atmaximum 750mm centres. Electricalconnections are achieved by attachingtight wire ties, integral to the anodesystem, to the steel reinforcement;and then the areas are instatedusing appropriate repair mortar.More recently, trials in Norway havedemonstrated that the woven carbonmesh with cementitious grout couldbe used as an anode system for thecathodic protection of reinforcedconcrete (Vennesland et al, 2006).

Not all of the anode systems,mentioned above proved effective,successful or suitable for all types ofstructural elements. The selection ofthe most suitable anode system(s)would depend on corrosionmorphologies and structural geometry.

Advantages of cathodic protection

The principal advantage of cathodicprotection over traditional repair isthat only damaged concrete areas

(i.e. spalled, delaminated or severelycracked) need to be replaced.Concrete, which is contaminatedwith chloride but otherwise sound,can remain since the possibilityof subsequent corrosion will beprevented by the appropriateelectrochemical process. The costsinvolved in the installation andoperation of the cathodic protectionsystem are more than offset bythe savings which result fromthe reduction in concrete repairquantities and shorter duration

of site work. In many cases, thereduction in repair may obviate theneed for temporary propping withconsequent reduction in costs.

A S S E T M

A N A G E M E N T

81 Corrosion mitigation of chloride-contaminated reinforcedconcrete structures: A state-of-the-art review


12/15611

Conclusions

The national and local government

policy of providing unrestrictedaccess to highway structures forfree movement of trade, commerceand other road users requires thatthey are well maintained. Theimportance of inspection and testing(using various NDT techniques foridentification and quantification ofdefects together with investigationinto the cause and the consequenceof the defects) was highlighted asan integral part of a well organisedprogramme for Highway MaintenancePlan. Without this information amanagement decision with regard totechnically correct and economicallycost effective repair/rehabilitationoptions may not be possible.

Various methods of protectingcorrosion damaged structures werediscussed and it was concludedthat cathodic protection is themost appropriate and proventechnique to stop corrosion of steelin reinforced concrete structures.

For long-term durability of repairof concrete structures damaged byreinforcement corrosion, particularlyin chloride-contaminated concrete,cathodic protection is recognised byhighways authorities and buildingowners as the most cost effectivemethod of concrete rehabilitation.The latest survey suggests that over1 million m2 of cathodic protectionsystems have been applied tohighway structures and buildingsworldwide. In the UK the applicationof cathodic protection systems hasbeen reported for over 200,000

m2 of concrete structures. A largenumber of installed CP systems are inoperation and have been performingsuccessfully for more than 20 years.

A S S E T MANA

GE ME NT

81Corrosion mitigation of chloride-contaminated reinforcedconcrete structures: A state-of-the-art review


13/15612

A S S E T M

A N A G E M E N T

81

References

ACI. Service–life prediction – state of art report, American Concrete Institute, Detroit, 2000, ACI committee 365.1.

Alexander, M.G. Concrete Repair, Rehabilitation and Retrofitting, Taylor and Francis, London, 2006.2.

Amey, S. L., Johnson, D. A., Miltenberger, M. A. & Farzam, H. Predicting the service life of concrete3.structures: an environmental methodology. ACI Structural Journal, 95, No. 2, 1998, 205-214.

BRITE/EURAM. The Residual Service Life of Reinforced Concrete Structures.4.Synthesis Report , Contract No. BREU-CT92-0501, 1995.

BSI British Standards. Cathodic Protection of Steel in Concrete, BSI, London, 2000, BS EN 12696.5.

Chess, P. M. Cathodic Protection of Steel in Concrete. E & N Spon, 1998.6.

Clifton, J. R. & Pommersheim, J. M. Predicting Remaining service l ife of concrete, Proceedings of7.International Conference on Corrosion and Corrosion Protection of Steel in Concrete, Sheffield, 1994.

Clifton, J. R. Predicting the service life of concrete. ACI Materials Journal, 90, No. 6, 1996, 611-617.8.

Cramer, S. D. Corrosion prevention and Rehabil itation Strategies for reinforced Concrete9.

Coastal Bridges. Cement and Concrete Composites, 24, No, 1, 2002, 101–117.Danish Standards Association. Repair of concrete structures to EN 1504: a guide for renovation of concrete10.structures, repair materials and systems to the EN 1504 series, Dansk Standard, Butterworth-Heinemann, 2004.

Das, S. C. Confidential Report; Cathodic Protection Design Document for Bridge in Essex, 1999.11.

Fraczek, J. A Review of Electrochemical Principles as Applied to Corrosion of Steel in a12.Concrete or Grout. Corrosion, Concrete, and Chlorides. ACI SP-102, 13-24, 2001.

Highways Agency. Cathodic Protection for use in Reinforced Concrete Highway Structures.13.Design Manual for Roads and Bridges, 2002, Volume 3, Section 3, Part 3, BA 83/02.

Koch, G. H., Brogers, P. H., Thompson, N., Virmani, Y. P. & Payer, J. H. Corrosion cost and preventive strategies in14.the United States. FHWA report, FHWA – RD – 01 -156, Federal Highway Administration, Washington DC, 2002.

NACE. Cathodic Protection of Steel in Atmospherically Exposed Concrete Structures, 2000, NACE RP0290.15.

National Cooperative Highway Research. Manual on Service Life of Corrosion-Damaged Reinforced Concrete16.

Bridge Superstructure Elements. Transportation Research Board of The National Academies, Washington, D.C.,2006

Neville, A. M. Properties of Concrete; Fourth and Final Edition, Pearson Education, 2002.17.

Tuutti, K. Corrosion of Steel in Concrete. Swedish Cement and Concrete Research Institute, Sweden, 1982.18.

Uhlig, H. H. Corrosion Handbook. John Wiley & Sons, 2nd Edition, 2000.19.

Vennesland, O., Hang, R. & Mork, J. H. Cathodic protection of reinforced concrete – a system with woven carbon20.mesh, Concrete Repair, Rehabilitation and Retrofitting, M. G. Alexander (eds), Taylor and Francis, London, 2006.

Wallbank, E. J. The performance of Concrete in Bridges: A survey of 200 Highway Bridges. HMSO, London, 1989.21.

Zhang, J. & Mailvaganam, P. Corrosion of concrete reinforcement and electrochemical factors in22.concrete patch repair. Canadian Journal of Civil Engineering, 33, No. 6, 2006, 785-793.



14/15613

Abstract

A S S E T MANA

GE ME NT

82Stochastic based lifecycle planningtool for ancillary pavement assets

Introduction

Transport Scotland is the agencyresponsible for the management ofthe trunk road network in Scotland.The recent Asset ManagementImprovement Programme (AMIP)undertaken by Transport Scotlandincluded a review of its existing assetinvestment planning capability1. Thisreview identified a number of areas ofcomparative weakness and, therefore,opportunities for the strengtheningof its investment planning capability.Foremost amongst these was aperceived weakness in the lifecycleanalysis capability for ancillary assets.

In the Transport Scotland context,ancillary assets encompass a widerange of asset types (includingroad markings, road signs, trafficsignals, safety barriers, etc.) asdefined in the Transport Scotland3rd Generation (3G) MaintenanceContract2. Historically, relatively

little effort has been expended oninvestment planning for ancillaryassets, when compared with (forexample) pavement and structuralassets. Whilst this situation isunderstandable given the respective

asset values and associated annualexpenditure, recent public spendingcuts in the UK have increased theneed for all trunk road agencies todemonstrate that scrutiny is beingapplied in all areas of expenditure.

The AMIP review of investmentplanning capability described above,made a number of recommendations

for improving Transport Scotland’sinvestment planning capability.Foremost amongst these was thedevelopment of a lifecycle planningtool for ancillary assets. Thedevelopment of any lifecycle planningtool typically requires a number ofkey elements to be in place, namely:

Inventory data – data on types•and quantities of assets

Condition data – current•asset condition

Intervention types – the range•

of treatments employed andtheir effect on asset condition

Unit rates – the costs of enacting•the various modelled treatments

Deterioration models – to•model the deteriorationasset condition over time

Scenarios - agreement on the•type of decisions that that toolwill be required to support

Modelling approach – a•modelling approach which isconsistent with all of the above

Implementation – of the•above in a software tool.

Because of the lack of historicprecedence in the modelling ofancillary assets in Transport Scotland,none of these necessary elementswere in place at the outset of theAMIP. A plan was therefore developedto put in place each of theseelements and thereby implementthe required investment planningcapability3. Whilst the above listimplies a particular order of work,

a number of tasks were executedin parallel. This paper outlines theprocess employed, and the variousdecisions made in the developmentof the ancillary asset lifecycle analysiscapability for Transport Scotland.

Highway AssetManagement Group

Atkins

Principal Consultant

Department of Civil andEnvironmental Engineering

The University of Auckland

Asset Management Branch

Trunk Roads NetworkManagement

Transport Scotland

Senior Lecturer

Senior Engineer

SeosamhB Costello

Angela JOwen

David CWightman This paper describes the application of a stochastic based lifecycle

planning tool, previously utilised for the strategic assessment ofpavement maintenance funding and policy decisions, to ancillarypavement assets such as road markings and safety barriers. The stochasticapproach recognises the fact that not all assets of a particular type willdeteriorate at the same rate, or in a strictly deterministic manner.

Historically, relatively little effort has been expended on investment planningfor ancillary assets, compared to the pavement and structures stock. Whilethis is understandable, given the relative asset values, recent public spendingcuts have highlighted the need to include all pavement assets in publicspending reviews. Unfortunately, because of this historic precedence, anunderstanding of current asset condition and future asset performanceis lacking in most authorities for their ancillary pavement assets.

Transport Scotland, the agency responsible for the management of alltrunk roads in Scotland, have addressed the former through the inclusionof condition assessment, as opposed to exception reporting, in theirdetailed inspection programme. A condition assessment manual has alsobeen developed to assist inspectors in carrying out their assessments.Knowledge of asset performance will come with time, however in theinterim, expert engineering knowledge has been harnessed to populatethe deterioration algorithms. The application of the analytical tool inTransport Scotland is demonstrated and lessons learnt reported.


15/15614

A S S E T M

A N A G E M E N T

82 Stochastic based lifecycle planning tool for ancillary pavement assets

would be conducted. Whilst objectivecondition measures can be employedfor certain assets (for example, retro-

reflectivity readings for road markingsand road traffic signs), for themajority of ancillary assets subjectivevisual assessment is required. Inorder to assist OC inspectors inmaking these subjective judgementswhen collecting condition data, acondition assessment manual wasdeveloped4. In an attempt to minimisethe subjectivity of inspections, themanual makes extensive use ofimages to illustrate asset condition,as well as detailed description of

assessment criteria. An extract fromthe condition assessment manual,relating to the condition of roadmarkings, is shown in Figure 1.

A key tenet of the assessmentmethod supported by the manualis the definition of asset condition(for all ancillary asset types) in termsof condition bands. Five conditionsbands have been adopted, namelyA, B, C, D and E, with A representing

Data collection

As described previously, a primary

barrier to the introduction of alifecycle analysis capability for ancillaryassets for many trunk road agenciesis a lack of knowledge regardingthe condition and performance ofsuch assets. This is something of acatch-22 situation. Since a modelcannot be developed without thiskey information, and conversely,there is little motivation to collectsuch data if it is not to be usedas part of a formalised decision-making process. A key phasetherefore in the development of theTransport Scotland capability was atargeted data collection exercise.

It should be noted at this pointthat whilst Transport Scotlandhas overall responsibility for themanagement of the Scottish trunkroad network, it subcontractsmany of its day-to-day, operationalresponsibilities to four OperatingCompanies (OCs), each havingresponsibility for a geographicalregion (i.e. South West, North West,North East, and South East).

The OCs were commissioned byTransport Scotland to undertakenthis data collection exercise. TheOCs were provided with data sheetson which 57 separate ancillaryasset types, as derived from the3G contract2, were listed. For eachof the listed asset types, the OCswere requested to provide valuesfor the following parameters:

Inventory – quantities for•each specified asset type (inthe appropriate units)

Asset lives – expected asset•lives / replacement intervals

Replacement costs – the unit•cost of replacing each asset type(accounting for regional variations)

Current condition – current•condition of assets.

Due to the historical precedentdescribed above, much of this datawas not readily available. In the caseof inventory data, asset quantitieswere obtained from the OCs’

Routine Maintenance ManagementSystem (RMMS) or from other localsources. Where such sources werenot available (or data was deemedto be unreliable), quantities wereobtained either via a limited surveyor by estimation. In the case of asset

lives, values were either derivedfrom the RMMS or other local datasources, or were based on expert

engineering judgement. Unit rates forasset replacement were derived eitherfrom OC framework rates or werebased on engineering experience.

Condition assessment

In the case of asset condition data,virtually no data was held by theOCs at the outset of the study. Thelack of condition data for ancillaryassets is a situation that is notunique to Transport Scotland. Themajority of RMMS, are not typically

configured to store asset condition,and inspection regimes typicallydo not cover asset condition. Thisis particularly the case for ancillaryassets. Furthermore, because of thediversity of the ancillary asset stockthere are no standardised approachesto assessing current condition.

In order to address this gap it wasdecided that a condition survey

Figure 1 - Extract from Condition Assessment Manual


16/15615

A S S E T MANA

GE ME NT

82Stochastic based lifecycle planning tool for ancillary pavement assets

new condition through to Erepresenting the end of the asset’sserviceable life. The condition of allassets is defined in terms of one ofthese five bands. The manual definesthe asset condition that constituteseach band. A field trial was conductedby one of the OCs to determine itssuitability for operational use. Thistrial identified a number of areas forimprovement, and a revised versionof the manual was then produced.Table 1 shows an extract of the data

collection return from one of the fourOCs. It can be seen in this table thatin addition to the total number ofeach asset type being specified, therespective OC has also specified (as apercentage) the distribution of assetsbetween the five condition bands.

It should be noted that the conditionassessment manual and the associatedcondition survey were not instigatedpurely for the purposes of this study. Itis the intention of Transport Scotlandthat the surveys will in future be

conducted at the intervals specified inthe manual. Amongst other benefits,this activity will provide a good baseof time-series asset performancedata which can be used in futureto calibrate asset performancemodels. Whilst the condition data

collection undertaken for this studyprimarily comprised desk-studiesand visual surveys, this does notpreclude the use of video surveysor other techniques in the future.

Model scenarios

At the outset of the study, TransportScotland specified that the newlifecycle planning tool should supportthe investigation of two primaryinvestment scenarios, namely:

Defined Condition – what•budget is required to maintainthe ancillary asset network ata defined condition level?

Defined Budget – what is the best•network condition that can beachieved for a specified budget?

On-going discussions withTransport Scotland as the projectprogressed identified twofurther investment scenarios:

Prioritisation 1 – where the•budget is insufficient to

achieve the defined condition,allocate funding to assets ona user-defined priority basis.What is the resulting impacton network condition?

Prioritisation 2 – where the•

budget is insufficient toachieve the defined condition,endeavour to ensure that user-defined minimum performance(condition) targets are met,before allocating the remainingbudget on a user-defined prioritybasis. What is the resultingimpact on network condition?

The ‘Prioritisation’ scenarios describedare a reflection of the financial climatein which many road agencies currentlyoperate. Whilst funding has not yet

reached the point inferred implied bythese scenarios, Transport Scotland,as responsible asset managers,are keen to investigate the likelyimpact should this scenario arise. Itshould be noted that whilst the toolresulting from this study is capable ofhandling each of the above scenarios,it is also capable of investigatinga wider range of scenarios, inaddition to those listed here.

Choice of modelling approach

A number of life-cycle planning tools,both deterministic and stochastic innature, are available to develop life-cycle plans. A variant of the stochasticmodelling tool developed by Costelloet al.5 has been used to develop life-

Table 1 - Extract from data collection return from selected Operating Company


17/15616

A S S E T M

A N A G E M E N T

Current condition

Current network condition is

represented in the model as adistribution of condition, defined asthe proportion of each asset grouplying in each condition band. Thisprovides the base year scenario orstarting point for life-cycle planning.

Mathematically, the initial state ofany process may be described by astarting vector, in which the sum ofall

i should be equal to one, and all

entries should be positive values. Thisstarting vector is represented by:

(1)In the Transport Scotland context,the Operating Companies havesupplied the condition distributionfor each asset type. as illustrated inTable 2 above. Taking the first assettype in this table as an example (i.e.Catchpit), it can been seen that atthe time of data collection, 16.75%of catchpits were adjudged to bein condition band A, 18.75% in B,34% in C, 17.75% in D and 12.75%in E. Using these percentages,

the starting vector for this assetgroup would be represented by:

(2)

In this way starting vectors for all228 asset group were derived.These starting vectors wereused to populate the model.

Deterioration modelling

In the stochastic-based approach,asset deterioration is modelled by

means of the transition probabilitymatrix (TPM), denoted by P. Thegeneral form of P is given by:

(3)

This transition probability matrixcontains all the information necessaryto model the deterioration of

the respective asset group. Thetransition probabilities, p

ij, indicate

the probability of the portion ofthe asset group in condition imoving to condition j in 1 year dueto the damaging effects of trafficand environment, as applicable.

cycle plans for a number of UK roadagencies, for both carriageway andstructure assets. Stochastic models

are also widely used elsewhere andtherefore provide a relatively well-proven approach. Prominent amongthese are the Transport ResearchLaboratory’s network conditionmodel6, the Finnish National RoadAdministration’s highway investmentprogramming system7 and thenetwork optimisation system8, asused by a number of state highwayadministrations in the USA.

The stochastic-based approachto lifecycle-planning typically

involves a number of common,fundamental concepts, as follows:

Asset Groups – the arrangement•of assets into groups of assetswith homogeneous characteristics(based on performance andreporting requirements)

Condition Bands – the•adoption of a single conditionmeasure which is rated interms of a discrete number ofcondition states or bands.

Condition Distribution – the•

range of condition states forthe various assets in each assetgroup is defined in terms ofa condition distribution.

Transition Matrices – the•deterioration of asset conditionis modelled using transitionprobability matrices whichmodel the transition of assetsfrom one condition bandto another, over time.

Each of these concepts is describedin more detail later in this paper.

Ancillary assets are deemed to behighly suited to the stochastic-based approach as follows:

Asset Groups - ancillary assets•naturally lend themselves tobeing modelled at a group leveldue to their diversity (the rangeof different asset types) and thehomogeneity within each group.

Condition Bands/Distribution - the•condition assessment approachadopted for this study (based onbands, and as embodied in theTrunk Road Condition AssessmentManual) naturally lends itselfto this form of modelling.

Transition Matrices – in the•context that there is negligibleempirical data on the performance

of ancillary assets, the transitionmatrix approach provides apragmatic means of establishing

initial deterioration models, basedon expert engineering judgement.

As asset performance databecomes available through thecondition monitoring programme,more sophisticated deteriorationmodels can be developed.

Model implementation

Having selected the modellingapproach to be employed, andhaving undertaken the data collectionexercise required to populate themodel, the next phase in the studywas the development of the actualmodel. The following sections describethe overall model build process, andin particular, how the various statedcomponents of the stochastic-basedapproach (namely asset groups,condition distribution, and transitionmatrices) were implemented.

Asset groups

Under the stochastic-based approachto life-cycle planning, individualassets are typically aggregated intoasset groups. These groups should behomogeneous in nature (particularlyin terms of performance) therebyallowing them to be modelled as agroup. At stated above, the TransportScotland 3rd Generation Contract2 defines the list of 57 ancillaryasset types to be managed by theOperating Companies. Given that thislist of asset types is well-established inthe Transport Scotland context, it wasdecided to retain this classification in

the lifecycle model. Further discussionsestablished that in addition toreporting model outcomes by assettype, Transport Scotland also wantedoutcomes to be reported by region (orOC). Therefore, it was decided thatthe model would employ 228 assetgroups (57 x 4). Whilst this numberis relatively high when comparedwith the number employed in otherimplementations of the stochasticapproach, such as that describedby Costello et al.5, the number was judged not to be excessive dueto a) the short model run-times,and b) flexible reporting facilitiesincluded in the model which allowthe end-user to decide the groupingsto be used in output reports.

Stochastic based lifecycle planning tool for ancillary pavement assets82


18/15617

A S S E T MANA

GE ME NT

Similar to the starting vector, forevery TPM the sum of the entriesin each row is equal to one, and

all entries are non-negative. Inmatrix notation, the probabilitydistribution at a specific elapsedtime in years, say t = 1, is given by:

(4)

Or more generally, the probabilitydistribution at any time tmay be calculated using:

(5)

Asset deterioration can thereforebe modelled using the aboveequation, where a

t is the distribution

of condition at time t, a0 the

distribution of condition at time0 (that is the starting vector), andPt the TPM raised to the powerof t, the elapsed time in years.

Two more conditions apply to theprocess when used to simulate assetdeterioration. First p

ij = 0 for i > j,

reflecting the general belief thatassets cannot improve in conditionwithout first receiving some form

of treatment. Secondly, pnn = 1,signifying a holding state wherebyassets that have reached theirworst condition cannot deterioratefurther. Consequently, in assetdeterioration the general form of thetransition matrix P is denoted by:

(6)

Initial deterioration models

Given that there was insufficient

historical data available to TransportScotland and its OC to be used inderiving performance models forancillary assets, expert engineeringopinion was instead used. Asdescribed above, estimates ofasset lives were captured duringthe data collection exercise. Aninteresting finding arising from thisexercise was the range of service lifeestimates for certain asset classesin the various regions. This findingreflects the very diverse nature ofthe Transport Scotland network,

ranging from heavily-traffickedurban environments in the southof the country, to relatively lowly-trafficked rural environments inthe north. This diversity in servicelives is also due to the diversity ofenvironmental conditions experiencedin the various regions, with weatherconditions often very severe innorthern regions (both in terms oftemperature and rainfall). It shouldbe noted that in estimating assetlives, it was assumed that routine

maintenance is undertaken on allassets at the specified intervals,for example, washing of sign-faces. Examples of estimatedasset lives are given in Table 1.

The estimates of asset life obtainedby this process were then used togenerate a separate TMP for eachasset group. For the time being, lineardeterioration rates were assumedfor all asset types. It should be notedthat the assumption of a lineardeterioration rate is purely a startingpoint from which to carry-out ananalysis in the first year of the newlyintroduced condition assessmentregime. As deterioration informationbecomes available through thecondition monitoring programme,these initial assumptions can bechallenged and the performancemodels refined accordingly.

The generic form of the transitionprobability matrix is given below,

where n is the number of condition

bands and L is the asset life assupplied for each asset type:

(7)

The generic form of the transitionprobability matrix is bestdemonstrated by means of anexample. Taking an imaginaryasset group as an example, wheren is 5 and L is 7, the resultingtransition probability matrixwould be as given below:

(8)

Using the above generic TPM form,and assuming a linear deteriorationrate, TPMs can be manually-derived.

Alternatively, the tool developedas part of this study has beenprovided with a facility to automatethe generation of TPMs for a user-specified L value (in the case ofthe Transport Scotland model, thevalue of n is fixed at 5). Figure3 shows a screenshot from therelevant part of the user interface.

Future refinements todeterioration models

Although the approach described

above enables transition probabilitymatrices to be determined withminimal data (in this case from just n and L values), the morecomprehensive the data-set used,the greater the confidence in theresulting performance models. Asactual asset lives and deteriorationrates become available through thecondition monitoring programmeinstigated as part of this study, theinitial transition probability matriceswill be refined accordingly.

Performance monitoring andcontinual feedback into theperformance models is necessary inorder to ensure that increased levelsof confidence in model forecastsare attained. The standard approachis to observe, from historical data,

Stochastic based lifecycle planning tool for ancillary pavement assets 82


19/15618

calculator developed by Costelloet al9. based on the methodology

defined by Ortiz-Garcia et al10

. Thetransition matrix calculator determinesthe probabilities in the transitionmatrices based on a deteriorationcurve, coupled with an estimate ofscatter or confidence in the model.

Analytical framework

The next phase in the study wasthe development of an analyticalframework in which the variousmodel elements established asdescribed above (asset groups,

starting condition vectors, andtransition probability matrices) couldbe encapsulated. In simple terms,the modelling process employedis based on two main elements:

i) the calculation of the annualprogression of conditiondistributions, followed by

ii) simulation of the effects ofrenewals applied in each year.

This iterative process is outlinedin Figure 4 and detailed in thesteps listed alongside it.

The deterioration modellingmechanism which is at the heart ofthis process can be best demonstratedby means of an example. Usingthe example given above, thedistribution of current condition forcatchpits reported above is taken asthe starting vector, a

0, and the TPM

for road markings: longitudinal (LL)developed above is taken as P. Then,to simulate 1 year’s degradation,the two matrices are multipliedtogether. This is represented below:

(10)

Each of the elements in the resulting

matrix, a1, is calculated by multiplyingthe first matrix, a

0, by each of the

columns in the second matrix inturn. The resulting distributionof condition, a

1, is as follows:

(11)

This means that after 1 year’sdeterioration 7% of the networkwill be in condition band A, 18% inB, 25% in C, 27% in D and 23%in E. This process can be continuedindefinitely to provide a distribution

of condition in future years.

Modelling of interventions

The scenario (demonstrating yearon year deterioration), assumesthat no maintenance interventionsare enacted. In practice renewalsare carried out on a yearly basisas dictated by the interventionlevels and available budget.

Assuming that catch-pits are renewedwhen they reach condition bandE, then 22.9% of the network willrequire renewal at the end of year1. When a renewal is applied it isassumed to restore the asset to itsoriginal condition. This is simulated byreturning the proportions renewed to

the way in which an asset groupdeteriorates over time and use this to

estimate pij using equation9

below. Nijis the number of assets in the assetgroup that moved from condition ito condition j during 1 year and N

i

is the total number of assets thatstarted the year in condition band i:

(9)

The proportions are likely to varyfrom year to year thereby requiring

an average to be determined overtime for each pij to ensure accuracy

in the model. Alternatively, asdeterioration curves are developedover time from historical data, thetransition probability matrices can bedetermined using the transition matrix A

S S E T M

A N A G E M E N T

Stochastic based lifecycle planning tool for ancillary pavement assets

Figure 3 - Facility for deriving TPM from L value

82


20/15619

condition band A. Assuming sufficientfunds are available, then 22.9% ofcatch-pits will be renewed at the

end of year 1, resulting in 30.1% ofthe network in condition band A,17.6% in B, 25.3% in C, 27% in Dand 0% in E. This then becomes thestarting vector for year 2, that is:

(12)

The budget required in year 1 toachieve this target performance canthen be calculated by multiplyingthe unit cost for catch-pits by thequantity then by 30.1%. The budget

required in subsequent years canbe obtained by repeating the aboveprocess as required. This essentiallyanswers the ‘what-if ’ question,‘what budget is required to supporta target network performance?’.

Budget allocations

The above example assumes thatsufficient funds are available torenew all catch-pits in band E.However, in practice this is rarelythe case. Consequently, renewal

budget limits can be specified withinthe model. This will also allowthe ‘what-if’ impact of differentlevels of funding on networkperformance to be modelled.

Initially, the annual budgets are set,followed by the onwards allocation of

the annual budgets to asset groups.Renewals are then applied untilthe relevant budget is exhausted,

assuming the required expenditureexceeds the available budget. Thisresults in a shortfall and parts of thenetwork remaining untreated. Thiscan be demonstrated by an example.

Assuming only a proportion of thefunds required are available, sayenough for only 3% of catch-pits,compared with the 22.9% requiringrenewal at the end of year 1, thenthis will result in 10.2% of catch-pits in condition band A, 17.6%in B, 25.3% in C, 27% in D and

19.9% in E. This then becomes thestarting vector for year 2, that is:

(13)

As before, the budget required insubsequent years can be obtainedby repeating the above process asrequired. However, if the availablebudget exceeds the budget requiredin a particular year, then a surplus inthe budget occurs. This surplus canthen be reallocated to a different

asset group. This can be carried outeither through a manual iterativeprocess whereby the user maintainscomplete control over the process orusing the automated capability builtinto Transport Scotland system.

A S S E T MANA

GE ME NT

Stochastic based lifecycle planning tool for ancillary pavement assets

Figure 4 - Analytical framework

82

1) The base year conditiondistributions or starting vectors, a

0,

are derived from current condition

data. Where inspection intervals arenot annual, as may be the case forasset types deemed to be low-risk,then the condition distributions

can be projected forward to thebase year, if required, using theappropriate TPM, to derive thebase year condition distributions.

2) The condition distribution ofeach asset group post deteriorationis predicted using the TPM.

3) The proportion of each asset group

requiring renewal is then calculatedbased on the post-deteriorationcondition distributions and thespecified intervention levels.

4) The percentage of eachasset group that will actually be

renewed is calculated based onthe specified budget constraints.

5) The effect that the renewalswill have on the condition ofeach asset group is calculated.

The above process is repeated for eachyear of the analysis period until theend of the analysis period is reached.

Conclusion

The stochastic-based lifecycle planning

methodology described in this paperhas shown itself to be appropriatefor the modelling of ancillary assetson the Transport Scotland network.Such assets naturally lend themselvesto being modelled at a grouplevel due to their homogeneity. Inaddition, the assessment criteriadefined in the condition assessmentmanual naturally lend themselvesto this form of modelling.

The approach as applied in this studyhas some limitations as it stands. The

deterioration models are based onengineering estimates of asset life andan assumption of linear deterioration.However, as deterioration informationbecomes available through thecondition monitoring programme,these initial assumptions can bechallenged and the performancemodels amended accordingly,thereby improving model accuracy.

Although subjective visual assessmentsare commonplace in inspections ofhighway assets, the reproducibilityof assessments conducted using thecondition assessment manual hasyet to be determined and furtherresearch is required. Consequently,until such time as actual deteriorationrates become available throughthe analysis of historical time-series data and the reproducibilityof the assessment method isdetermined, the results of the modelshould be used with caution.

Acknowledgement

The study on which this paper isbased was funded by the ScottishGovernment as part of WorkPackage 12 of the Transport ScotlandAsset Management ImprovementProgramme (AMIP). The authorswould like to acknowledge thesupport of all staff involved inthis study (whether directly orindirectly) both from TransportScotland itself and from within its’Operating Companies. In particularwe would like to acknowledge thesupport of the following individuals:

David Arran and Willie Grant(of Transport Scotland), and BillMoss, Derek McMullen and ChrisKrechowiecki-Shaw (of Atkins Ltd).


21/15620

A S S E T M

A N A G E M E N T

References

Atkins. Review of Road Management Process, Work Package 12: Specification of a Computerised1.

Decision Support Tool, Asset Management Improvement Programme, Transport Scotland, 2008.Management, Inspection And Cyclic Maintenance, 3rd Generation Term Contract for Management and2.Maintenance of the Scottish Trunk Road Network - North West Unit, Schedule 7 Part 1, Transport Scotland, 2005.

Atkins. Specification and Performance Models for Ancillary Assets, Work Package 12: Specification of a3.Computerised Decision Support Tool, Asset Management Improvement Programme, Transport Scotland, 2009.

Transport Scotland. Transport Scotland Trunk Road Condition Manual, Asset4.Management Improvement Programme, Transport Scotland, 2010.

Costello, S. B., M. S. Snaith, H. G. R. Kerali, V. T. Tachtsi, and J. J. Ortiz-Garcia. Stochastic model for strategic5.assessment of road maintenance. Proceedings of the Institution of Civil Engineers, Transport 158(4): 203–211.

Kerali, H. G. R., and M. S. Snaith. NETCOM: The TRL Visual Condition Model for Road Networks.6.Transport Research Laboratory, Crowthorne, UK, Contractor Report 321, 1992.

Thompson, P.D., L. A. Nuemann, M. Miettinen, and A. Talvitie. A micro-computer Markov dynamic7.

programming system for pavement management in Finland. Proceedings of the Second NorthAmerican Conference on Managing Pavements, Toronto, Vol. 2. pp. 2.241–2.252, 1987.

Kulkarni, R. B. Dynamic decision model for a pavement management system. In Transportation8.Research Record: Journal of the Transportation Research Board, No. 997, TransportationResearch Board of the National Academies, Washington, D.C., 1984, pp. 11-18.

Ortiz-Garcia, J. J., S. B. Costello, and M. S. Snaith. Derivation of transition probability matrices for9.pavement modelling. Journal of Transportation Engineering ASCE 132(2): 141–161, 2006.

Costello, S.B., J. J. Ortiz-Garcia and M. S. Snaith. Development of the transition10.matrix calculator. In Proceedings of the Tenth International Conference on Civil,Structural and Environmental Engineering Computing, Rome, Italy, 2005.

Stochastic based lifecycle planning tool for ancillary pavement assets82


22/15621

Water & Environment

Atkins

Senior LandscapeArchitect

C A

R B ON C R I T I C AL DE S I GN

83

Paul Fraser

Carbon Critical Masterplanning

Introduction

The BIM revolution in architecturestarted in the 1980s, redefiningthe process of designing a buildingenvelope through the use of anintegrated design data model.By attaching metadata to drawnobjects, it became possible to movebeyond simple digital drafting toa dynamic design package thatgenerates schedules, visualisationand procurement information in realtime. The scope of these packages isoften thought of as ‘5D’ - the three

traditional dimensions of design,plus time (4D) and cost (5D).

Masterplanners have over timemade limited use of architecturalBIM packages for their own ends,occasionally building bespoketools, but there is very littlecurrently on the market by way ofparametric capability specificallyfor masterplanning. Althoughmasterplanning and architectureboth rely on interrelationshipsbetween drawn data and

analytical/reporting tools, theparameters are very different.

Successful masterplans, see Figures1 and 2, balance physical form with

a broad range of factors that willultimately determine its successas a place. In parallel with theinvestigation of BIM style solutionsfor our masterplanning businessby the London studio, the brief

was being developed under AtkinsCarbon Tools programme for a

tool to enable the forecasting ofthe carbon profile of a masterplan.The decision was taken to join thetwo workstreams together, andHolistic City Ltd were commissionedto build a bespoke Atkins carbontool for their CityCAD platform.

The scale and pace of global urbanisation is truly remarkable. A recentreport estimates that India alone needs to build the equivalent of a newChicago every year until 2030 to meet their demand, investing over $1trillionin the process 1. Urban change on this scale coupled with the need todecarbonise the global economy challenges both clients and professionalsto take a holistic approach to sustainability, from the fundamentalsof land use planning to building typologies and infrastructure.

Under Atkins Carbon Tools programme, a team from across the MasterplanningTechnical Network has developed a platform that will help us meet thesechallenges. A parametric ‘BIM-style’ (Building Information Modelling)platform has been developed that enables all data variables embodiedin a masterplan to be tracked and optimised in real time - includingboth embodied and operational carbon. This allows the design teamto understand the complex interrelationships between factors affectingsustainability outcomes in urban systems at the level of a masterplan. Thispaper explains the core capabilities of the tool, how it will revolutionise ourapproach to masterplanning, and the benefits of its use for our clients.

Abstract

Figure 1 - Suhar Urban Masterplan (Oman): City centre development proposals


23/15622

C A R B O N C R I T I C A L D E S I G

83 Carbon Critical Masterplanning

How do you calculate the carbonimpacts of a Masterplan?

The tool allows the design team toarrive at aggregated values for bothembodied and operational carbonacross a major scheme of up to 50kmx 50km (depending on the complexityof modelling). It is possible to usethe tool in a detailed way, modellingcollections of buildings with ancillaryspaces, or in a more strategic mode –analysing representative developmentmixes in super blocks. This lattermethod makes allowances for, butdoes not actually model, 3D buildings.

Software architecture

For the tool to be globally applicable,

it required a centrally hosted databasefor the carbon factors and the use ofSequel Server facilitated connectionto desktop machines. The databasewas developed by a specialist thirdparty supplier, Accelero Digital, withthe data specification devised by theSustainability team to make the finalcalculations as robust as possible inthe context of the masterplan scale ofdesign. Integration with the CityCADapplication required the developmentof two new components to the core

platform - one to facilitate retrieval

of information from the database,and a second to define the changesto the user interface, including thenew calculations and outputs. Theapplication testing was done by a joint UK/China team. Developmentof the tool is now complete andthe London, China and India teamsare discussing its application witha range of potential and existingclients. Licences are available throughthe Atkins Carbon Tools Portal2.

Figure 2 - White City, Baku (Azerbaijan): Illustrative Masterplan


24/15623

83

In order for the carbon calculationto be meaningful, the outputs arebroken down by four key themes -

land use, transport, waste treatmentand renewable energy, for bothoperational and embodied carbon.This enables the team to see wheredesign decisions have the mostimpact, see Figure 3, (either positiveor negative), understanding theimpact of design changes in realtime. The software works by pairingassumptions (or factors) providedby the database with actual datameasured from the 3D model. Anexample might be the potential for a

building to accommodate electricity-generating technologies. Takingphotovoltaic panels as an example,the database provides assumptionson what type of technology wouldbe appropriate, the percentageof roof area available for it, itsmounting angle etc. The modelprovides the available roof area,and the two are used to calculatethe predicted electrical output.This is done for each building, withfurther positive contributions fromstand alone generation kit (wind

turbines etc.) to build a profile for thewhole masterplan. In parallel withgeneration, the model also considerselectricity demand, allowing theMasterplanning team to understandthe net energy demand/generationof the scheme, and the resultingload on the local grid supply. Thisis one example of a whole range ofmetrics that can be used to ‘balance’the masterplan, making it trulysustainable. The tool also includesa spatial trip distribution module to

analyse carbon emissions associatedwith transport. This module considersthe relative locations of key trafficgenerating and attracting land usesto determine a value for transportoperational carbon. It incorporatesa range of variables to enable theTransport Planning team to calibratethe model to its context, and allowsthe design team to ‘test’ the relativelocations of key land uses such ashospitals and schools. Also importantfor the fundamental efficiency ofthe masterplan, in carbon terms, is

the building orientation and blockefficiency. The tool enables the user tosee instantly the relative efficiency andresulting density of different blockforms, see Figure 4, combined withother factors such as the resulting

waste generated. The databasesupplies a value for the optimumorientation of a given building typein its geographic location, and themodel provides measured quantities(for instance the actual roof area ororientation of a block. These factorsare then combined in the model’sdatabase, combining across themasterplan to give a complete pictureof the development’s properties.Because the database uses a relativelysimple (yet powerful) method of

using parameters as multipliers oncore values, such as Gross Floor Area(GFA), it is capable of tracking ahost of numbers that are based onthe development mix – much like atraditional methodology. One exampleof this is its ability to generate acost model for the developmentby allowing the team to input costper square metre construction costsalongside income projections toarrive at an initial understanding ofthe margins likely to be achieved

Figure 4 - Heat mapped view for anlysis of key factors

Figure 3 - Diagram showing the importance of land use decisions for sustainability


C A

R B ON C R I T I C AL DE S I GN


25/15624

83

and the resulting viability thresholdsof the site. This means we have thecapacity to build in the client’s cost

model from the outset, working withthem to optimise the development’slayout including cost considerations- rather than it being measured bythe client (or his Quantity Surveyor)after the design has been completed.

Benefits for our clients

The parametric capability of theCarbon Critical Masterplanningtool gives the Masterplanning teamthe capability to navigate complexinterrelationships between citymetrics in real time. This is the keyto the power of the system. Thequestions of the impacts of relativeactivities in a proposed city couldhave been answered before, butthrough separate workstreams andwith time delays while informationwas processed by different teams.The masterplanning carbon toolmitigates these possible delays toprovide a more holistic solution.

It also places the decarbonisationof the masterplan at the heart

of the design process, see Figure5, meaning that both the designteam and the technical specialismsinvolved can see the impact of designdecisions on the masterplan’s carbonfootprint. By being highly visual, isalso makes these factors accessibleto a wide range of lay stakeholders,including the client, members of thepublic, local authority officers etc.

Alongside carbon optimisation,typical scenarios of interestto clients might include:

What will be the impact on the(1)financial model of the use ofenhanced building standards?

Can I optimise the phasing(2)and distribution of floorspaceagainst the capacity of thelocal property market? Whatwill the impact be in terms ofnew infrastructure delivery?

What social infrastructure(3)will I need to provide for thenew population, and whenwill it need to be built?

How can we optimise the(4)energy use of the developmentto minimise the load on thelocal utilities networks?

What will the population(5)

of this new city be? Howmany jobs will it create?

How much waste will the(6)development produce? Whathappens to the energy demandsif we handle some of it on site?

Figure 5 - Parametric design process using Carbon Critical Masterplanning



Summary

Atkins' Carbon Tools were developedto place carbon optimisation at theheart of our business. By integratingthe Carbon Critical Masterplanningtool with a BIM style masterplanningplatform, we are able to offer carbonoptimisation as a cost effective partof our core process, rather than a

bolt on requiring parallel resource.Although the implementation ofthe tool is in its early stages, initialfeedback from our internationaldesign teams has been positive, and itis being piloted on live projects. Alsofundamental to its success is its abilityto enable smarter and more efficientworking, allowing the masterplanningteam to analyse design iterationsmore quickly and ultimately delivera better service to our clients.

References Mckinsey Global Institute (2001) India’s urban awakening: Building inclusive cities, sustaining economic growth1.

2. http://www.holisticcity.co.uk


26/15625

Business Manager

Highways & Transportation

Atkins

84

Steven Fraser

Mitigating transport's climate change impactin Scotland: Assessment of policy options

Introduction

The Scottish Governmentand climate change

The Scottish Government (SG)published its Government EconomicStrategy in 2007. This states thatthe purpose of the SG is to:

“focus the Government and publicservices on creating a more successfulcountry, with opportunities for allof Scotland to flourish, throughincreasing sustainable economicgrowth” (The GovernmentEconomic Strategy, p1)24.

In support of the Strategy, the ClimateChange (Scotland) Act received

Royal Assent on August 4, 2009. TheAct sets in statute the GovernmentEconomic Strategy target to reduceScotland's emissions of greenhousegases by 80% by 2050, one of thesustainability purpose targets8.

This covers the basket of sixgreenhouse gases (as recognisedby the United Nations FrameworkConvention on Climate Change),and includes Scotland's share ofemissions from international aviationand international shipping.

The Act also establishes an interimtarget for 2020 of at least 42%reductions in emissions, and allowsMinisters, by order, to vary thereduction figure for the interimtarget based on expert advice fromthe advisory body. Progress towardsthese targets will be driven by aframework of annual targets.

The Act is a key commitment of theScottish Government, and is the mostfar-reaching environmental legislationconsidered by the Parliament duringthe first ten years of devolution.

The bold targets signify theimportance Scotland places onplaying its part in mitigating oneof the most serious threats facingour world. Figure 1 graphicallyillustrates the scale of this challenge.

Transport’s contributionto climate change

In addition to the wider Scottishtargets referenced above,reducing emissions from transportspecifically is also one of the SG’sNational Transport Strategy’sthree key strategic outcomes.

In 2007, Scottish transport, includinginternational aviation and shipping,accounted for 14.7 mega-tonnesof carbon dioxide equivalent(MtCO

2e), or 25.9% of total Scottish

greenhouse gas emissions.

Reducing emissions from transport is one of the Scottish National TransportStrategy’s three key strategic outcomes. On December 5 th 2008 the ScottishGovernment published the Climate Change (Scotland) Bill, which includesa commitment to reduce emissions by 50% by 2030, and by 80% by2050. The finalised Act also includes an interim target of a 42% reductionby 2020. These targets demonstrate a bold commitment by the ScottishGovernment. It signifies the importance Scotland places on playing itspart in mitigating one of the most serious threats facing our world.

The Scottish Government’s Transport Directorate wanted to improve its evidencebased on how it can contribute to meeting emission reduction targets andappointed Atkins, in partnership with the University of Aberdeen, to undertakea study to identify, analyse and report on the policy options available tothe Scottish Government. The analysis involved an assessment of individualpolicy options (including the ultimate production of marginal abatement costcurves including each policy) and also an assessment of central and ambitiousscenarios that were formed by packaging together complementary policies.

The study has identified 22 devolved policy options that are available to theScottish Government. A number of broad patterns have emerged in relationto the relative performance of different types of policy options. The analysissuggests that the Car Demand Management (Smart Measures) category hasthe greatest potential to reduce CO

2 emissions. The fiscal policy options also

offer significant abatement potential, however, the analysis suggests mostof the infrastructure policy options would offer significantly less potential.

The model results suggest that the combined effect of the policy options in theCentral Scenario would achieve an annual abatement of around 1.35MtCO2 p.a.

in 2022, whilst the Ambitious Scenario would achieve an additional 0.80MtCO2,

representing a total of 2.15Mt CO2 p.a. in 2022. These represent an 8% and

12% reduction respectively against projected transport emissions for that year.

Abstract


27/15626


84 Mitigating transport's climate change impact in Scotland:Assessment of policy options

Study methodology

Key objective:The key objective of the studywas to identify, analyse and reporton the devolved policy optionsavailable to mitigate transport’sclimate change impact in Scotland.

Seven stages

The study took eight months tocomplete and was undertakenusing a seven stage methodology:

Establishment of a preliminary(i)list of potential policy options

220527412 geotechnics for the structural engineer

Documents