effects of climate change on road subgrades muzi

EFFECTS OF CLIMATE CHANGE ON ROAD SUBGRADES

by

MUZI BONGINHLANHLA MNDAWE

Submitted in partial fulfilment of the requirement for the degree

MAGISTER TECHNOLOGIAE: CIVIL ENGINEERING

in the

Department of Civil Engineering

FACULTY OF ENGINEERING AND THE BUILT ENVIRONMENT

TSHWANE UNIVERSITY OF TECHNOLOGY

Supervisor: Prof JM Ndambuki

Co-Supervisor: Dr WK Kupolati

November 2014

ii

DECLARATION BY CANDIDATE

I hereby declare that the dissertation submitted for the degree M Tech: Civil

Engineering, at Tshwane University of Technology is my own original work and

has not previously been submitted to any other institution of higher education. I

further declare that all the sources cited or quoted are indicated and

acknowledged by means of a comprehensive list of references.

MUZI BONGINHLANHLA MNDAWE

Copyright © Tshwane University of Technology 2014

iii

ACKNOWLEDGEMENTS

The research was for a Masters Technology dissertation partly sponsored by

Royal Haskoning DHV whom I would like to thank for their support with funding,

study leave and helpful advice. Their kindness in affording me the opportunity to

utilise the facilities of Soilco Materials Investigations (PTY) LTD at their expense

helped improve cost and time savings during this study.

I greatly appreciate the guidance, assistance and advice rendered by my

academic supervisors, Prof Julius Ndambuki and Dr Williams Kupolati who have

helped to improve my research. Their prompt efforts and attention in this study is

highly appreciated. Dr Adedayo Badejo, a postdoctoral fellow at the Department of

Civil Engineering, Tshwane University of Technology, is also appreciated for his

guidance and support during the final stages of the research.

I would like to express my sincerest gratitude for the constant inspiration and

critical evaluation of the study by my colleague, Mr Robbie Dunbar. Lastly, I would

like to thank my wife, Lerato Mndawe for her constant support, patience and

prayers.

iv

ABSTRACT

Climatic data is one of the most important inputs required in any road design. The

historical information being currently relied upon for pavement design may soon

lose its significance due to the expected global climate change. The broad

objective of the study was to determine and simulate future climate change for

road sub-grades in the southern African region, with a view to developing new

pavement design parameters in order to protect the pavement infrastructure.

The methodology included gathering information from archives of past

observations and future simulated weather trends managed by the Council for

Scientific and Industrial Research (CSIR) in South Africa and the Commonwealth

Scientific and Industrial Research Organisation (CSIRO) in Australia. It also

included extensive soil laboratory testing, particularly California Bearing Ratio

(CBR) tests on identical samples that were compressed after soaking at varying

daily intervals.

Results from the analyses indicated an increase in magnitude of extreme daily

rainfall events with return periods of 10 to 30 years in the entire southern Africa.

Furthermore, the findings revealed an equation ideal for use on weaker gravels to

prominently reduce the turnaround time from 5 to 2 days when determining the

CBR. The research also showed that when the CBR of the soaked pavement fell

to 23% of its four day strength, the pavement carrying capacities declined by 51%

of their original value. This will subsequently change the pavement categories and

correspondingly affect the design reliability of pavements.

v

TABLE OF CONTENTS

PAGE

DECLARATION BY CANDIDATE ii

ACKNOWLEDGEMENTS iii

ABSTRACT iv

LIST OF FIGURES viii

LIST OF TABLES ix

GLOSSARY x

CHAPTER 1: INTRODUCTION .............................................................................. 1

1.1 BACKGROUND ............................................................................................ 1

1.2 PROBLEM STATEMENT .............................................................................. 3

1.3 RESEARCH OBJECTIVE ............................................................................. 4

1.4 RESEARCH SIGNIFICANCE ........................................................................ 5

1.5 SCOPE ......................................................................................................... 8

1.6 BRIEF METHODOLOGY .............................................................................. 8

1.7 LAYOUT OF DISSERTATION ...................................................................... 9

1.8 STUDY AREA ............................................................................................. 10

CHAPTER 2: LITERATURE REVIEW .................................................................. 12

2.1 ROAD SUBGRADE ..................................................................................... 12

2.2 SOUTH AFRICA’S ROAD NETWORK ........................................................ 12

2.3 CLIMATE CHANGE EVIDENCE ................................................................. 14

2.4 FACTORS AFFECTING PAVEMENT DESIGN .......................................... 15

2.4.1 Sub-grade strength and material selection ............................................... 17

vi

2.4.2 Extreme temperature variations ............................................................... 18

2.4.3 Rainfall and moisture in sub-grades ......................................................... 19

2.4.4 Population growth and urbanisation ......................................................... 21

2.5 FUTURE CLIMATE SIMULATIONS ............................................................ 22

2.6 ENVIRONMENTALLY OPTIMISED DESIGNS ........................................... 22

2.7 REPAIR COSTS DUE TO CLIMATE CHANGE .......................................... 24

CHAPTER 3: MATERIALS AND METHODS ........................................................ 25

3.1 FIELD VISITS.............................................................................................. 26

3.2 CLIMATE DATA COLLECTION .................................................................. 26

3.3 SOIL SAMPLE DATA COLLECTION .......................................................... 29

3.3.1 Determining the degree of accuracy required .......................................... 30

3.3.2 Determining the sampling frequency ........................................................ 31

3.4 DATA ANALYSIS TOOLS ........................................................................... 32

CHAPTER 4: RESULTS AND DISCUSSION ........................................................ 34

4.1 PAST AND SIMULATED RAINFALL ........................................................... 35

4.2 DECADAL CHANGE IN EXTREME RAINFALL .......................................... 41

4.3 SOIL LABORATORY TESTS AND RESULTS ............................................ 45

4.3.1 SOILCO site laboratory results................................................................. 48

4.3.2 Graphical presentation of SOILCO site laboratory CBRs ......................... 53

4.4 PROJECTED CLIMATE CHANGE .............................................................. 56

4.5 EFFECT OF EXTREME RAINFALL ON SOUTHERN AFRICAN REGION . 56

4.6 EFFECT OF MOISTURE ON CBR TEST ................................................... 57

4.7 INTERPRETATION OF STRENGTH AND FORMULAE USAGE ............... 59

4.8 RESULTS WITH REFERENCE TO OBJECT OF INVESTIGATION ........... 61

4.9 ASSUMED ELASTIC MODULI FOR VARIATION IN CBR AND MOISTURE

CONTENT ......................................................................................................... 62

4.10 REQUIRED DESIGN LIFE ........................................................................ 63

vii

4.11 SIMULATED PAVEMENT CARRYING CAPACITY IN RELATION TO SUB-

GRADE MOISTURE ......................................................................................... 64

CHAPTER 5: CONCLUSION AND RECOMMENDATIONS ................................. 67

5.1 CONCLUSION ............................................................................................ 67

5.1.1 Macro climatic regions of southern Africa ................................................ 67

5.1.2 Rapid assessment of CBR ....................................................................... 68

5.1.3 Reduced pavement carrying capacity ...................................................... 69

5.2 RECOMMENDATIONS ............................................................................... 70

REFERENCES ..................................................................................................... 72

APPENDICES ...................................................................................................... 79

viii

LIST OF FIGURES

PAGE

Figure 1.1: Macro Climatic Regions of Southern Africa .......................................... 5

Figure 1.2: Thornthwaite Moisture Index ................................................................ 6

Figure 1.3: Southern Africa and uMkhanyakude locality map............................... 10

Figure 2.1 Sources of moisture in pavement ........................................................ 20

Figure 2.2 Implications for choice of road surfacings ........................................... 23

Figure 4.1 Monthly mean rainfall (mm) Pongola Experiment Station Farm 1967 -

2001 ..................................................................................................................... 39

Figure 4.2 Annual mean rainfall ............................................................................ 39

Figure 4.3 Monthly mean rainfall .......................................................................... 40

Figure 4.4 Annual mean rainfall ............................................................................ 40

Figure 4.5: Macro Climatic Regions of southern Africa ........................................ 42

Figure 4.6: Decadal change in extreme rainfall .................................................... 43

Figure 4.7: Projected change in the annual frequency of extreme rainfall events 44

Figure 4.8 Variation of CBR with time of soaking ................................................. 54

Figure 4.9 Variation of CBR with time of soaking ................................................. 55

ix

LIST OF TABLES

PAGE

Table 1.1 Comparison of Weinert and Thornthwaite climatic indices ..................... 7

Table 2.1 extent of South Africa’s road network ................................................... 13

Table 4.1 Monthly rainfall (mm); Pongola Experiment Station Farm1967 – 2001. 35

Table 4.2 Annual monthly rainfall (mm) at Pongola Experiment Station Farm 1967

– 2001 .................................................................................................................. 36

Table 4.3 Monthly rainfall (mm) at Pongola Experiment Station Farm 2046 – 2065

............................................................................................................................. 37

Table 4.4 Annual rainfall (mm) Pongola Experiment Station Farm 2046 – 2065 .. 38

Table 4.5 RoadP443/1 compacted at varying moisture contents ......................... 46

Table 4.6 RoadP435/1 compacted at varying moisture contents ......................... 47

Table 4.7 Road P443/1 soaked at varying durations ............................................ 49



Table 4.10 Road P435/1 soaked at varying durations .......................................... 52

Table 4.11 Road P443/1actual and predicted CBR values .................................. 54

Table 4.12 Road P435/1 actual and predicted CBR values ................................. 55

Table 4.13: Road P443/1 actual and predicted CBR values (composite equation)

............................................................................................................................. 60

Table 4.14: Road P435/1 actual and predicted CBR values (composite equation)

............................................................................................................................. 60

Table 4.15: P443 and P435/1 pavement design ................................................... 61

Table 4.16: Moduli Values v CBR Values @ 90% MAASHTO density ................. 63

Table 4.17: Traffic Data Available ........................................................................ 63

Table 4.18: Simulated pavement carrying capacity for variation in moisture

content; P443 ....................................................................................................... 65

Table 4.19: Simulated pavement carrying capacity for variation in moisture content

for P435 ................................................................................................................ 66

x

GLOSSARY

Alluvium Loose, unconsolidated soil or sediments, which has

been eroded, reshaped by water in some form, and re-

deposited in a non-marine setting

Bearing Capacity The amount of mass that soil can hold without it giving

way

Bedrock The solid rock that underlies loose material, such as

soil, sand, clay, or gravel.

Calcrete A conglomerate of surficial gravel and sand cemented

by calcium carbonate

Climatology The scientific study of climate

Capillary The ascension of liquids through slim tube, cylinder or

permeable substance due to adhesive and cohesive

forces interacting between the liquid and the surface

Cretaceous sediments a naturally occurring material that is broken down by

processes of weathering and erosion, having the

quantities of chalk ,clay, silt, sand, and gravel, mostly

of non-marine and near shore marine origin

Embankment A mound of earth built to hold back water or to support

a roadway or railway

Glacial retreat A condition occurring when backward melting at the

front of a glacier takes place at a rate exceeding

forward motion

Global circulation model A class of computer-driven models for weather

forecasting, understanding climate and projecting

climate change

Groundwater The water located beneath the earth's surface in soil

pore spaces and in the fractures of rock formations

Highway A main road or thoroughfare, such as a street,

boulevard, or parkway, available to the public for use

for travel or transportation

xi

N-value An expression of the climatic area or a region based on

the weathering and durability of natural road building

materials

Pavement Part of a roadway having a constructed surface for the

facilitation of vehicular movement.

Resilient modulus A mathematical description of an object or substance's

tendency to be deformed elastically when a force is

applied to it

Roadbed The material at which the road embankment is to be

constructed

Rutting A depression of the pavement in the wheel path

Salination The process of increasing the salt content in soil

Seal Consists of a coat of bituminous binder sprayed into

the road surface which is then covered with a layer of

aggregate, stone or sand

Storm surges An abnormal rise in the level of the sea along a coast

caused by the onshore winds of a severe cyclone

Sub-base A layer in a pavement system between the subgrade

and base course or between the subgrade and the

concrete pavement

Sub-grades The native material underneath a constructed sub-base

Undulations Wavelike motions visible on the pavement surfacing

usually associated with settlement of embankments

Water table The level below which the ground is saturated with water

Weatherability The property of a material that permits it to endure or

resist exposure to the weather

1

CHAPTER 1: INTRODUCTION

1.1 BACKGROUND

When dealing with transportation infrastructure, climatic data is one of the most

important inputs required in any road design. The available historical information

which has until now been relied upon will soon be of less importance due to the

expected global climate change. Past and current seal designs and bearing

capacity of soils have been based on this climatic data, which in effect suggests

that as climate data changes, seals and bearing capacity requirements will change

too. Thus, adaptation is an essential part of the response to the threat of climate

change.

Engineers, particularly in southern Africa are facing the challenge of utilising old

methodologies for pavement designs while the roads are designed for a lifespan of

between two and three decades. When taking changes in climatic conditions,

increasing vehicle loads and traffic volumes into consideration, the pavement

system is bound to fail and an aggressive approach to counter the plight of climate

change in the transportation sector is urgently required. This simply means that

regardless of any adaptation measures that may be developed for pavements, it

may be too late for any remedial measures to be effective if implemented on the

already constructed infrastructure. For example, most construction projects take

place on soil and fewer projects are carried out on solid bedrock. Therefore the

ability of soil to support mass must be evaluated prior to the construction of any

pavement. The mass that a soil can hold without it giving way is called its bearing

2

capacity. A soil's bearing capacity will vary depending on what type it is and is

affected by environmental factors external to properties of the soil.

So far, research carried out on climate change has primarily focused on

highlighting the impacts on the road network and quantifying the cost of

maintaining or restoring the existing infrastructure. The United States of America

(USA), Scotland and Australia are the only countries known to have significantly

conducted research on climate change impacts on road infrastructure. However,

their research is limited and it does not deal with the effects on sub-grades

(Chinowsky et al., 2011). Their findings merely detail predicted trends in climate

change. The main gap in the research was the exclusion of the methods of

incorporation of climate change into the design of road pavements.

According to Youman (2007), current design practices will soon, no longer be

adequate for road and infrastructure assets to cope with the anticipated effects of

climate change. For example, higher water tables can accelerate the rate of

pavement deterioration due to capillary action increasing the moisture content of

pavements. Road agencies may need to raise the levels of existing embankments

when pavements reach the ends of their useful lives. The design of new roads

should therefore provide for anticipated rises in water tables in susceptible areas

such as coastal roads to avoid saturation and complete failure of the sub-grade

through capillary movement of groundwater (Impact of climate change on road

infrastructure, 2004). In South Africa, more attention with regards to climate

change has been focused on coastal roads with emphasis on rising sea levels and

salinity. Although climate data is of importance in pavement design, current design

3

methods have not yet made provisions for future climate change. Climate change

effects will have a negative impact on road infrastructure in the near future and

necessary developments should be established in order to protect such

infrastructure. According to Trademark Southern Africa (2011), a study initiated by

the African Development Bank is currently underway which addresses the effect of

climate change on the road infrastructure on the African continent as a whole.

Existing literature related to climate change adaptation in the infrastructure sector

is primarily qualitative in nature with an emphasis on broad recommendations and

warnings.

In addition to this, the influence of field variables such as density and

environmentally related parameters such as temperature and equilibrium moisture

content of the pavement layers on the Mechanistic Empirical design input and

models were largely unquantified (Theyse et al., 2007).

This study investigated one aspect of climate change which is rainfall. The rainfall

measurements making up the past data were recorded from weather observation

stations near the study area. The simulation of the effect of rainfall was then

conducted through the CBR test using subgrade material quality.

1.2 PROBLEM STATEMENT

Climate change poses an imminent threat to all mankind as such, protection of our

road infrastructure from its effects is essential. There are currently general

assumptions being made with respect to climate change and its effects on the road

infrastructure, particularly sub-grades. These assumptions suggest that rutting and

4

failure of the pavement structure will in future occur more frequently and that

reconstruction together with maintenance programs will be required earlier in the

design life of the pavement than initially anticipated (Mndawe et al., 2013). It can

therefore be inferred that incorporating future climate change effects and possible

scenarios in pavement designs is essential and pavement designs should be

carried out considering the impacts of climate change.

1.3 RESEARCH OBJECTIVES

The main objective of the study was to determine the impacts of climate change

on road sub-grades, with a view to developing new pavement design parameters

in order to protect the road infrastructure. In achieving this objective, particular

attention was given to the future relevance of Weinert N-values as depicted in the

Macro Climatic Regional Map of southern Africa adopted from Weinert (1980) and

extended soaking of soils using the California Bearing Ratio (CBR) test which is a

measure of the strength of granular materials.

The specific objectives of the research were:

i. to analyse extreme rainfall effects on southern African region

ii. to determine the required bearing capacity of soils for road construction

projects under changing weather patterns

iii. to develop and integrate an equation for shorter determination of CBR of

weaker gravels.

5

iv. to simulate the potential damage caused by future climate change effects in

current sub-grades

1.4 RESEARCH SIGNIFICANCE

Extensive research has been conducted by climatologists on the subject of climate

change. However, engineers and technologists have not yet adopted an approach

that aims to address the topic within the engineering arena. Improvements ought

to be made particularly on climate and time based parameters used in

transportation engineering and designs.

Figure 1.1 shows the Macro Climatic Regional Map of southern Africa adopted

from Weinert (1980) by Technical Recommendations for Highways (TRH4) (1996).

It is one of the most outdated weather based catalogues used in the industry. To

date, even in light of the imminent threat of climate change, no credible advances

have been made yet for any improvements on this over thirty year old design

climatic regional map.

Figure 1.1: Macro Climatic Regions of Southern Africa

Source: Technical Recommendations for Highways (1996)

6

There is also a close relationship between the Weinert N-Value as expressed in

the Macro Climatic Regions Map and the Thornthwaite Moisture Index. The

Thornthwaite moisture index, which is a function of rainfall, temperature and

potential evapotranspiration, is also considered for modelling purposes on roads

projects. Roads in areas with higher value for the Thornthwaite index will

deteriorate faster than those with a lower value for the same traffic loading. Figure

1.2 shows the Thornthwaite Moisture Index map.

Figure 1.2: Thornthwaite Moisture Index


In this research, the climatic data obtained was used to revise only the Weinert N-

Value map. The difference between Weinert N-Values and Thornthwaite Moisture

Index is that where N is less than 5 in the former, rocks are likely to decompose as

the value suggests the area may be moderate to wet. The latter also gives an

indication of the overall availability of moisture during the year using a different

7

formula where a wet region may have an index figure of up to 100 as shown in

Table 1.1

Table 1.1 Comparison of Weinert and Thornthwaite climatic indices

Description Weinert N value

Thornthwaite

Moisture Index,

Im

Typical Mean

Annual Rainfall

(mm)

Arid 5+ < -40 < 250

Semi-arid 4 to 5 -20 to -40 250 to 500

Semi-arid to sub-

tropical

2 to 4 0 to 20 500 to 1000

Humid tropical < 2 20 to 100 > 1000

Source: TRL Limited (S.a:15)

Furthermore, the California Bearing Ratio (CBR) test method has been used for

more than seven decades with very limited improvements in its lifetime especially

with regards to the time it takes to conduct the test. It is considered one of the

most fundamental tests of any granular material in road construction.

It takes any soil laboratory a period of at least seven days to produce a

comprehensive set of CBR and Indicator tests. The former is in essence a five day

long test method and can be completed on time if commenced on a particular day

of the week; that is Monday, Thursday and Friday. The waiting period means that

whatever progress can be made with regards to construction on site will in the

meantime be based only on experience of site technical staff and very little

8

scientific input. Therefore there is a need to make improvements on current test

methods in order to expedite such a lengthy test procedure.

1.5 SCOPE

The study involves an improvement of two fundamental components of pavement

designs particularly in the southern African region. It attempts to revise the Macro

Climatic Regions of southern Africa and the CBR method as conducted in soil

laboratories. These two components are integral tools for any pavement designer

considering the role that climate plays in pavement design and not to mention the

time lost in waiting for soil laboratories to complete one of the most fundamental

tests required in any road construction site. The study then estimates the reduced

carrying capacity due to climate change based on the extended soaking of

specimens used for subgrade construction. The extended soaking is a simulation

of future wet spells on subgrade quality materials.

1.6 BRIEF METHODOLOGY

The study included extensive soil laboratory testing, particularly CBR on identical

samples that were compacted at variable moisture contents which are above, at

and below Optimum Moisture Content (OMC) and then compressed after soaking

for various periods. The results of such tests for all the specimens were then

plotted to obtain a trend that best represents the data such that a formula could

also be developed. An equation aimed at obtaining the CBR of materials within a

shorter timeframe than the current five day period it takes to compact, soak and

9

compress the soil specimen was then derived from the data (See Appendices E, F

and G).

The study also involved the application of a variable-resolution atmospheric global

circulation model, the Conformal-Cubic Atmospheric Model (CCAM) of the

Commonwealth Scientific and Industrial Research Organisation (CSIRO) in

Australia (Piketh et al., 2012) to obtain an ensemble of six regional climate

projections that are analysed as part of this research.

1.7 LAYOUT OF DISSERTATION

This dissertation is divided into five chapters. Chapter one provides the

introduction and also presents the background to the study, problems that

prompted the research, the objectives, significance, scope of the study and

methodological approach of the study.

Chapter two presents a literature review. In this chapter, both the theory and the

approaches of climate change and its impacts are reviewed. Chapter three

presents the detailed methodology used in the study including laboratory testing,

data collection, analysis and presentation of the data obtained.

Chapter four deals with the results obtained from climate data simulations as well

as the detailed laboratory test. Thereafter, the chapter presents the discussion and

interpretation of the results. Finally, chapter five summarises the conclusions of

the investigation and outlines recommendations and suggestions for further study.

10

1.8 STUDY AREA

The study was conducted in two segments with the first one focusing on the entire

Southern African region. The second segment focuses on two adjacent roads

located within uMkhanyakude District Municipality namely P444 in Jozini and P435

in Ndumo. They are located in Northern KwaZulu-Natal and border Swaziland and

Mozambique. The study area is bound in the west by the Lubombo mountain

range, which reaches an elevation of approximately 600m above sea level.

Figure 1.3: Southern Africa and uMkhanyakude locality map

Source: Digital Map Studio, (2014)

11

The area is characterised by seasonal dry winters and wet summers with periodic

flooding. Typical high rainfall season is November to January and the average

annual rainfall is in excess of 1200mm while periodic flooding normally occurs

during the later periods of the annual summer rainfall season. The summer

temperature ranges from 23°C to 40°C, while winter temperatures range from

16°C to 25°C. Soil along the Lebombo Range consists mainly of shallow, stony

soils of the Mispah and Glenrosa forms formed from the lavas of the Lebombo

volcanics. The soils found along the floodplain and in particular along the west

bank of the Pongola River, are derived alluvium, river terraces and the Cretaceous

sediments.

12

CHAPTER 2: LITERATURE REVIEW

2.1 ROAD SUBGRADE

The sub-grade is that portion of the earth roadbed which after having been

constructed to reasonably close conformance with the lines, grades, and cross-

sections indicated on the plans, receives the selected, subbase, base and surface

layers. In a fill section, the sub-grade is the top of the embankment or the fill. In a

cut section the sub-grade is the bottom of the cut. The sub-grade supports the

sub-base and/or the pavement section. According to Schaefer (2008), the

performance of a pavement depends on the quality of its sub-grade and sub-base

layers. As the foundation for the pavement’s upper layers, the sub-grade and sub-

base help mitigate the detrimental effects of climate and the static and dynamic

stresses generated by traffic.

According to Davies (2004), the resilient modulus of the underlying material

supporting the pavement is now considered as a key material property in the

mechanistic-empirical design procedures. Attempts have been made by

researchers to predict the sub-grade resilient modulus from laboratory and field

experimental methods based on the soil properties

2.2 SOUTH AFRICA’S ROAD NETWORK

According to Kannemeyer (2010), South Africa’s total road network is about

747 000 kilometres, of which over 154 000 km are paved or surfaced roads. The

South African National Road Agency (SANRAL) is responsible for the country's

network of national roads, which grew to over 20 000 km with an estimated value

13

of over R40-billion in 2010. About 3 000 km of the national roads are toll roads.

About 1 800km of these are maintained by SANRAL, while the rest have been

concessioned to private companies to develop, operate and maintain. Table 2.1

depicts the extent of South African Road Network.

Table 2.1 Extent of South Africa’s road network

Authority Paved (km) Gravel (km) Total (km)

SANRAL 16 170 0 16 170

Provinces – 9 48 176 136 640 184 816

Metros – 9 51 682 14 461 66 143

Municipalities 37 691 302 158 339 849

Total 153 719 453 259 606 978

*Un-Proclaimed (Estimate) 140 000 140 000

Estimated Total 153 719 593 259 746 978

*Un-Proclaimed Roads = Public roads not formally maintained by any Authority

Source: Kannemeyer, (2010)

SANRAL asserts that there is currently a R50 billion backlog on strategic (national

and provincial) roads, with an associated maintenance budget of R12 billion

annually (Sabita 2011). As noted in South Africa and internationally, road

maintenance delayed for one year increases repair costs to between three and six

times (SAICE Infrastructure Report Card for South Africa, 2011).

According to The World Bank (S.a), proper road maintenance contributes to

reliable transport at reduced cost, as there is a direct link between road condition

14

and vehicle operating costs (VOC). An improperly maintained road can also

represent an increased safety hazard to the user, leading to more accidents, with

their associated human and property costs.

2.3 CLIMATE CHANGE EVIDENCE

According to Fairhurst (2008), changes in water-tables (i.e. elevation resulting

from sea level rise and salt water intrusion) have the potential to cause serious

engineering problems in developing and built up areas. Moreover, sea level rise is

one of several lines of evidence that support the view that the climate has recently

warmed (Cartwright, 2008).

One of the best-documented evidence of climate change is the glacial retreats that

have been witnessed on Mount Kilimanjaro in Africa. It is the tallest peak on the

continent, and despite being located within the tropics, it is high enough that

glacial ice has been present for at least many centuries. However, over the past

century, the volume of Mount Kilimanjaro’s glacial ice has decreased by about

80% (See Appendix A). If this rate of loss continues, its glaciers will likely

disappear within the next decade. Similar glacial melt backs are occurring in

Alaska, the Himalayas, and the Andes (Climate Institute, 2012).

According to the National State of The Environment Report-South Africa (S.a),

there are four stations recording sea level rise on the coast of Southern Africa.

These are located at Lüderitz (Namibia) and in South Africa at Port Nolloth,

Simon's Town and Mossel Bay. Records of the first three stations indicate a

15

positive trend in sea-level rise (relative to the land mass) over the past three

decades. For example, the trend at Port Nolloth is a 12.3 mm rise per decade.

Since the west coast sea-level rise data are in agreement with the global trends, it

is reasonable to accept that the predicted rates of sea level rise, modelled on the

basis of global warming, are applicable to South Africa.

2.4 FACTORS AFFECTING PAVEMENT DESIGN

The Department of Environmental Affairs and Tourism (2005) states that in terms

of the impacts of climate change on South Africa, recent studies predict that

climate change will cause mean temperature increases in the range of between

1°C and 3°C by the mid-21st century, with the highest increases being witnessed

in the most arid parts of the country. A broad reduction of rainfall in the range 5% -

10% has been predicted for the summer rainfall season. This is likely to be

accompanied by increased incidences of both drought and floods, with prolonged

dry spells being followed by intense storms. A marginal increase in early winter

rainfall is predicted for the winter rainfall region of the country. A rise in sea level

by as much as 0.9m by 2100 is also predicted (Department of Environmental

Affairs and Tourism, 2005).

According to the Li et al., (2011), current highways are designed based on typical

historic climatic patterns, reflecting local climate and incorporating assumptions

about a reasonable range of temperatures and rainfall levels. Given anticipated

climate changes and the inherent uncertainty associated with such changes, a

pavement could be subjected to very different climatic conditions over the design

16

life and its design might be inadequate to withstand future climate forces that

impose stresses beyond environmental factors currently considered in the design

process.

For example, flexible pavements under the same conditions are often affected by

bleeding, weathering, undulations, rutting, potholes and longitudinal and

transverse cracking. Some of these distresses are formed in combination with

traffic loads and or material defects. If extreme climate changes were to occur,

these distresses will clearly be exacerbated and new distresses may be formed (Li

et al., 2011), (See Appendix B).

The main impacts on road infrastructure may come from changes in flood heights

and sea level rise with storm surges (Austroads, 2004). According to Koch (2011),

sea-level is now rising faster along the U.S. Atlantic coast than at any time in the

past 2100 years, and this surge is linked to increasing global temperatures. The

anticipated effects of climate change should be manageable with current

engineering practice and the materials available, possibly with adaptation.

However, to provide certainty that our road network and transport infrastructure

can be properly managed, there needs to be more understanding on the impacts

of climate change on infrastructure (Youman, 2007).

The influence of field variables such as density and environmentally related

parameters such as temperature and equilibrium moisture content of pavement

layers on the South African Mechanistic Design Method input and models are

largely unquantified (Theyse, et al., 2007).

17

2.4.1 Sub-grade strength and material selection

According to the Bureau for Industrial Co-operation (2011), the strength of road

sub-grade is commonly assessed in terms of the California Bearing Ratio (CBR)

and this is dependent on the type of soil, its density, and its moisture content.

Where a mechanistic design approach using linear elastic theory is employed for

flexible pavements, the measure of sub-grade support is commonly assessed in

terms of the elastic parameters (modulus, Poisson’s ratio). The following factors

must be considered in determining the design strength/stiffness of the sub-grade:

i. The compaction moisture content and field density specified for

construction

ii. Moisture changes during service life

iii. Sub-grade variability

iv. The presence of weak layers below the design sub-grade level

A combination of moisture, density, CBR and swell which will give the greatest

CBR and density consistent with an acceptable amount of swell must be selected.

The CBR and density values so selected are those which must be considered in

the design of overlying layer thickness.

The selection of materials for pavement design is based on a combination of

availability, economic factors and previous experience (Guidelines for Human

Settlement Planning and Design, 2009). The strength of granular materials is often

susceptible to water, and excessive deformation may occur when water enters

through surface cracks. The water susceptibility of a material depends on factors

18

such as grading, the PI of the fines, and density (Guidelines for Human Settlement

Planning and Design, 2009).

2.4.2 Extreme temperature variations

Temperature affects ageing of bitumen through oxidisation and embrittlement

leading to cracking which permits infiltration of water through the cracked surfacing

and thereby weakening the sub-grade. (Impact of Climate Change on Road

Infrastructure, 2004). When temperature increases, the result is often a decreased

pavement stiffness which affects load distribution. According to the Washington

Asphalt Pavement Association (2012), pavements like all other materials, will

expand as they rise in temperature and contract as they fall in temperature. Small

amounts of expansion and contraction are typically accommodated without

excessive damage. However, extreme temperature variations can lead to

catastrophic failures.

Flexible pavements can suffer large transverse cracks as a result of excessive

contraction in cold weather. These cracks will increase the water content in the

unbound layers thereby reducing the bearing and load distribution capacity of the

layer. Rigid pavements on the other hand are prone to slab buckling as a result of

excessive expansion in hot weather. These can result in an excessively high

maintenance cost and total failure of the road infrastructure as a whole if solutions

for climate change effects on sub-grades are not developed soon enough.

19

2.4.3 Rainfall and moisture in sub-grades

Climate change can have direct and indirect impacts on road infrastructure. The

direct impacts are primarily due to the effects of moisture fluctuations, which

weaken flexible pavements, rendering them more susceptible to damage by heavy

vehicles and shortening their lives. It is believed that about 80% of road distresses

and pavement damages are related to the presence of excess water, water that

affects the behaviour of all layers, bound asphaltic material layers, granular layers

and sub-grades (Carerra et al., 2010).

According to Pavement Age (2012), moisture can significantly weaken the support

strength of natural gravel materials, especially the sub-grade. This is because

moisture can enter the pavement structure through cracks and holes in the surface

due to rainfall, laterally through the sub-grade, and from the underlying water table

through capillary action (See Figure 2.1). The result of moisture ingress is the

lubrication of particles, loss of particle interlock and subsequent particle

displacement resulting in pavement failure. A greater intensity of storm rain causes

increased watercourse flooding and consequent damage. Rising water on the

pavement may block traffic and cause damage to road equipment, and softens the

pavement structure with increased risk of damage and shortened lifetime (Norem

and Moller, 2007).

20

Figure 2.1 Sources of moisture in pavement

Source: Schaefer, (2008)

Increased rainfall in frequency and intensity results in higher ground water level.

Climate change in the form of rainfall storms is expected to have an impact on the

effectiveness of sub-surface drainage, especially if the amount of rainfall will

exceed the design capacity of the drainage system.

According to Technical Recommendations for Highways 4 (1996), experience has

also shown that inadequate drainage is probably responsible for more pavement

distress in Southern Africa than inadequate structural or material design. As a

21

result, effective drainage is essential for good pavement performance, and it is

assumed in the structural design procedure.

According to Youman (2007), granular materials are the predominant pavement

material for the lower pavement layers. These materials perform poorly under the

effects of water and are likely to be vulnerable to rising water tables or water levels

in coastal areas corresponding to sea level rise. Frequent cycles of wetting and

drying will also limit performance of granular pavement layers. Both of these

impacts are likely to occur in many areas due to climate change.

2.4.4 Population growth and urbanisation

According to Impact of Austroads (2004), current population projections show the

world’s population may exceed 15 billion by the end of the century. The

consequent demands on the production and consumption of goods and services

and for land, energy and materials will greatly intensify pressure on the

environment and living resources throughout the world. This change in population

further impacts on the natural flow of storm water as urbanisation will be

consequently increased.

Although not directly related to climate change, urbanisation is also another factor

that aggravates flooding. It restricts where flood waters can go, covering large

parts of the ground with roofs, roads and pavements. This obstructs sections of

natural channels and building drains that ensure that water moves to rivers faster

than it did under natural conditions (Mendel, 2006).

22

2.5 FUTURE CLIMATE SIMULATIONS

According to Climatology and Climate Change (2009), simulations indicate that

future temperatures will increase and rainfall patterns will fluctuate depending on

the particular season of the year. Summer rainfall (December to February) is

expected to remain similar to present day with possible small increases of about

20% or 15 to 25 mm. The early winter season (March to May) could result in a

continuum of present quantities of rainfall with possible slight (20 – 30%)

decreases. During winter, (June to August), there is likely to be either very slight

increases of between 5 and 10 mm in the mean monthly rainfall or a continuum of

current rainfall conditions. The rainfall simulations for spring (September to

November) show a mixed picture, ranging from an approximate 40 – 80%

decrease in rainfall in September to a possible 40% increase in November.

2.6 ENVIRONMENTALLY OPTIMISED DESIGNS

Fifty per cent of the damage reported is attributable to environmental deterioration.

Recent Department for International Development (DFID) and Swedish

International Development Cooperation Agency (Sida) funded projects in southern

Africa indicate that environmental factors other than climate alone are more

significant than traffic in influencing the performance of low volume sealed roads in

the tropics. According to Rolt et al., (2004) the environmental contribution is

believed to be as great as the traffic contribution on low volume rural roads. As a

result, an environmental factor should ideally be introduced in order for

deterioration models to cope with the range of climates envisaged. Thus one of the

principal reasons why engineers have not taken full advantage of the opportunities

23

for reducing the cost or for improving the quality of roads where traffic does little

damage is simply lack of knowledge or confidence to design the roads specifically

for particular environments. The implications of this are shown in Figure 2.2.

Where adequate design considerations to mitigate environmental factors are

adopted, low volume sealed roads can be provided at a low cost and as an

attractive alternative to gravel roads considering problems relating to their

sustainability and riding quality.

Figure 2.2 Implications for choice of road surfacing

Source: Rolt (2004)

24

2.7 REPAIR COSTS DUE TO CLIMATE CHANGE

According to Chinowsky et al., (2011), African continent is facing the potential of a

US$183.6 billion liability to repair and maintain roads damaged from temperature

and rainfall changes related to climate change through the year 2100. As detailed

by Chinowsky et al. (2011), the central part of the continent faces the greatest

impact from climate change with countries facing an average cost of US$22 million

annually, if a proactive adaptation policy is adopted. However, if a reactive

approach is adopted the costs will likely be US$54 million annual average,

(Chinowsky et al., 2011). This evidence therefore suggests that urgent proactive

measures in the form of improving current design parameters are required in order

to curb the impacts of climate change on sub-grades.

25

CHAPTER 3: MATERIALS AND METHODS

The research was chronologically conducted to meet the stated objectives. The

popularity of the study area elicited comprehensive and detailed information

gathering. Climatic data proved to be the hardest to collect as it is a specialist field.

Furthermore, the dissemination of the climate data required that it be processed by

the Council for Scientific and Industrial Research (CSIR).

Material data were obtained from reports of recently completed and ongoing

projects which were commissioned by the Department of Transport and managed

by Royal Haskoning DHV within the study area. At the time there were four

construction projects running simultaneously within a radius of 30km. The projects

were for roads P522/2, D9, P443/1 and P435/1. Two soil laboratories were used

during the study. The Tshwane University of Technology soil laboratory in Pretoria

was used for conducting the initial tests and thereafter a SOILCO Proprietary

Limited (Pty) Ltd site laboratory located at the centre of the four projects was used.

The KZN Department of Transport provided the most recent traffic data through its

website. As there was only one station in the study area, the traffic data from the

station was utilised for the other routes as well. Most of the data processing for the

research was then done from first principles. This included the design traffic and

pavement carrying capacity estimations.

26

3.1 FIELD VISITS

The weather observation stations used for collecting the data were Makhathini and

Phongola, which are located 15 kilometres North East of Jozini and within a 50

Kilometre radius respectively. Field visits were undertaken to ascertain the location

of the research stations, data collection process and clarify any misinterpretations

of the data. All aspects of the data collection process were explained on the day of

the visit at the Makhathini Research Centre.

A second field visit to the actual study area by the University supervisors took

place on Friday 26 October 2012 (See Appendix C). This was to ascertain the

sampling area, methods and other procedures that were required to be adhered

to. Other places visited included the CSIR which partly provided the climatic data

used in this research.

The Tshwane University of Technology materials laboratory in Pretoria was visited

and utilised in November 2012. All of the initial tests were conducted in this

laboratory: however, SOILCO materials laboratory became the laboratory of

choice due to distance from the sites on which the study is based. This laboratory

was used during January 2013.

3.2 CLIMATE DATA COLLECTION

The first step toward building future climate data was collecting a detailed data of

past and predicted weather. The data gathered were for air temperature and

27

rainfall. Data were gathered for the periods which ranged from 1940 to 2001 and

forecasted for the period 2012 to 2062. This data were obtained from the following

sources:

i. Climate Studies, Modelling and Environmental Health for the Council for

Scientific and Industrial Research (CSIR)

The data was maintained by the CSIR and reported a number of variables.

However, the data format was in Network Common Data Form (NetCDF)

which could not be decoded with the available computer software and an

alternative source that would avail the data in plain text was sought.

ii. Climate Systems Analysis Group (CSAG)

These data were maintained by the University of Cape Town’s (UCT)

Climate Systems Analysis Group which also report a variety of data and the

data format is in comma-separated values (CSV) file which can be imported

by an ordinary Microsoft Excel spreadsheet program. This format was user-

friendly and gave a comprehensive view of the required climate data.

iii. Conformal-Cubic Atmospheric Model (CCAM)

The data is maintained by the Commonwealth Scientific and Industrial

Research Organisation (CSIRO) in Australia and was used to obtain the

ensemble of six regional climate projections for Southern Africa that are

analysed as part of this research. This data was easily manipulated in order

to produce a basis for a reviewed Climatic Regional map of southern Africa.

28

Other sources of such data include the South African Risk and Vulnerability Atlas

(SARVA) whose intent is providing up to date climate information for various

sectors.

The methodology adopted in this study also included identifying and mapping

areas within the southern African region that may suffer from future increased

rainfall and flash flooding among other climate based phenomena.

A systematic approach was used in the process that entailed:

i. Desk study

ii. Data gathering

iii. Compilation of map

The Weinert N-Value, initially developed by Weinert (1974) and improved in 1980

was adopted by Technical Recommendations for Highways (TRH) 4 (1996) and

TRH 14 (1985). It was originally developed in southern Africa to describe different

climatic environments with respect to weathering and decomposition of rocks. It is

calculated using 12 times the evaporation (Ej) of the warmest month (January)

divided by the total annual rainfall (Pa) as shown in the formula;

a

j

P

EN 12

where;

N = Weinert N-Value

Ej = Evaporation

Pa = Annual rainfall

29

The data actually needed for computation of the N-value are annual average

rainfall, and evaporation rate. Using the N-value formula, contoured maps of N-

value were then developed for the Southern African region on a decadal basis

from 1961 – 2061. These contoured maps are an indication of the natural

weatherability of rock which is used as a regional adjustment criterion for all

pavement designs. This currently implies that in a wet climatic region where N < 2,

decomposition rather than disintegration would be the predominant mode of

weathering of the material.

3.3 SOIL SAMPLE DATA COLLECTION

The samples were taken in a random manner. Randomness, however, did not

imply that samples were taken haphazardly. A random sample is usually taken

according to a set of random numbers. In order for a random sample to be taken,

the road is divided into sample units. For practical purposes, the procedure

described below was used;

i. The procedure is described in detail in Method MC1 of TMH5, (S.a) and was

used for the sampling process. Test samples were taken from sections of

roads P443/1 and P435/1 that were already under construction. The soil

samples were taken from a depth of over 0.44m to 1.00m below the

surfacing. The tests were staggered as outer wheel track/inner wheel track

one side, outer wheel track/inner wheel track other side, centreline, etc.

ii. A total of eight samples extracted from each of the roads were seen as

adequate to provide data for the analysis considering it was sampled from a

30

uniform section on the road. This was gathered from the soil profiling

conducted during the design phase of the project.

iii. Two extra samples were taken from a nearby borrow pit with selected

material already in use for sub-grade and other selected layers on road

P435/1 to Ndumo Game Reserve. Sampling from borrow pits containing

sub-grade material (G7-G10) was carried out as this material had been

previously assessed visibly and seemed to conform to sub-grade material of

G10 quality . This G10 material was also already being used on route

P435/1 as a sub-grade layer.

3.3.1 Determining on the degree of accuracy required

The research required that a few soil samples be collected out of a large number

and its main characteristics are representative of the entire route or study area.

Several routes were identified for the sampling process. These routes were P522,

D9, P435 and P443/1 which are all within 50km from the study area’s main town

Jozini (Figure 1). Road D9 was reported to have been built on a calcrete sub-

grade thereby making this material unsuitable for the study in terms of material

classification as calcretes are strong, durable and sometimes water sensitive

Construction on road P522/2 was already complete together with asphalt paving

and further exploration in terms of sampling would have required disturbance of

the completed road pavement structure. With layer works construction still

underway on P435/1 and P443/1, these routes became the preferred choice as it

would also eliminate further establishment costs while promoting ease of sampling

31

and testing due to the currently constructed layer works and proximity of the site

soil laboratory.

A non-probabilistic but convenient sampling method was used in this research that

is, the sampling routes from the study area were pre-selected in a non-random

manner due to their convenience. This convenience determining factor was mainly

due to time constraints, the proximity of the routes to the study area and the fact

that these routes were currently under reconstruction thereby reducing red-tape

with regards to obtaining sampling permission from the Department of Transport.

One of the reasons which resulted in the choice of the sampling method employed

was that already constructed roads would have sub-grades more relevant to the

study than borrow pit material. It should also be noted that most construction

projects are now either partial or complete reconstruction which in either case, the

sub-grade material often undergoes minimal changes.

3.3.2 Determining the sampling frequency

The pavement and materials investigations were performed according to the

methodology described in Technical Methods for Highways (TMH) 5 document,

Method MC1. Every effort was made to collect all samples during one field

mobilization, requiring approximately 2 days of sampling and air drying by a team

consisting of soils laboratory manager, researcher and two laboratory assistants.

The investigations consisted of the following:

32

i. Visual inspection of the road condition and materials along the road. This

was done to confirm the limits of uniform pavement sections as detailed in

the soil profiles prior to the design stage. It was also conducted to identify

localised areas of potential problem materials and areas with drainage

problems, which could influence the performance of the pavement and

consequently the research sampling.

ii. Test pits were excavated at various pre-selected positions along the width

of the existing road including some of the localised previously identified

problem areas.

3.4 DATA ANALYSIS TOOLS

Initially, all the weather data analysed was received from the CSAG. Raw data

was also available from the CSIR website, however, with very limited usage as the

main requirement for these data was for it to be translated into a high-resolution

regional projection map of Southern Africa. A variable-resolution atmospheric

global circulation model, the conformal-cubic atmospheric model (CCAM) of the

Commonwealth Scientific and Industrial Research Organisation (CSIRO) in

Australia, was used to obtain the ensemble of regional climate projections that

were analysed as part of this project.

Soil laboratory data was prepared by manual calculation and further analysed

using Microsoft Office Excel 2007 for graphs and other visual supporting

information.

33

The empirical method described in TRH4 (1996) was used for the calculation of

the remaining pavement carrying capacity in the later phases of the study. The

KZNDoT website was visited to obtain traffic data along this portion of road.

Station Number 2489A is the only station on P443 and occurs at the eastern end

of the road at the P522 intersection. No other more recent counts were available.

34

CHAPTER 4: RESULTS AND DISCUSSION

This chapter presents and discusses the results obtained from climate simulations

for the past and the future. The data graphically shows the changes to be

anticipated in future in terms of rainfall. The climate data has been used to further

forecast extreme weather conditions for future reference. It is expected that in the

next century, climate will change in ways as difficult to imagine as it was in the turn

of the twentieth century when comparing with the present.

The chapter further discusses laboratory test results, also presented in detail in the

previous chapter. Although the laboratory tests were conducted in two phases and

at separate laboratories, the results presented still met the objectives set out at the

beginning of the research. The use of the equation developed for the quicker

determination of the CBR strength of subgrade material is discussed in

comparison with the conventional testing method.

Finally, the results are used to simulate pavement carrying capacity in relation to

subgrade moisture. This takes into consideration the variation of CBR strengths

with extended soaking periods as described in the materials and methodology

chapter. Real data is used as obtained from the roads on which the study is

focused.

35

4.1 PAST AND FORECASTED RAINFALL

The climate data presented represents a combination of past and forecasted

rainfall for the latter phase of the desired period of interest being the years 2012 to

2065 (Table 4.1 to 4.4). These data, albeit on a micro scale, are further used to

depict a true graphical presentation of the findings (Figure 4.1 to 4.4).

Table 4.1 and 4.2 represent monthly and annual mean monthly rainfall data

obtained from the CSAG. Owing to the non-availability of reliable data from the

weather stations adjacent to the study area, the closest station where the most

reliable data was obtained from was the Pongola Experiment Station Farm. The

station ID is 410144.1 and is located at -27.4° latitude and 31.58° longitude and is

within a 75km radius from the study area.

Table 4.1 Monthly mean rainfall (mm) at Pongola Experiment Station Farm

1967 – 2001

MONTH MEAN

Jan 111

Feb 97

Mar 80

Apr 40

May 21

Jun 11

Jul 8

Aug 14

Sep 31

Oct 75

Nov 109

Dec 101

36

Table 4.2 Annual mean monthly rainfall (mm) at Pongola Experiment Station

Farm 1967 – 2001

Table 4.3 and 4.4 present the forecasted mean monthly and annual mean monthly

rainfall data obtained from the CSAG for the years 2046 – 2065. The tables are

PERCENTILE

YEAR MEAN 10th 90th

1967 52 0.8 119

1968 29 1.4 64

1969 64 1.3 124

1970 36 12.3 69

1971 63 3.0 130

1972 48 3.8 103

1973 68 1.2 143

1974 60 2.2 152

1975 65 4.5 147

1976 49 0.0 111

1977 51 2.7 123

1978 67 10.9 161

1979 45 6.2 83

1980 58 0.8 154

1981 47 5.0 70

1982 38 0.3 81

1983 54 6.4 102

1984 88 19.2 107

1985 48 1.2 91

1986 50 3.0 120

1987 57 4.4 126

1988 62 14.1 122

1989 85 9.6 194

1990 38 2.6 74

1991 72 11.1 189

1992 37 0.0 93

1993 58 9.3 151

1994 50 2.2 133

1995 61 1.7 169

1996 73 4.6 150

1997 90 37.2 166

1998 56 0.8 128

1999 59 12.3 110

2000 124 21.1 213

2001 37 0.0 74

37

based on data obtained from the Pongola Experiment Station Farm. The station ID

is 410144.1 and is located at -27.4° latitude and 31.58° longitude and is within a

75 km radius from the study area. The data are all shown graphically in Figures

4.1 to 4.4.

Table 4.3 Monthly rainfall (mm) at Pongola Experiment Station Farm 2046 –

2065

PERCENTILE

MONTH MEAN 10th 90th

Jan 87 66 112

Feb 92 76 118

Mar 88 63 112

Apr 68 46 87

May 39 28 53

Jun 21 9 36

Jul 26 15 33

Aug 36 24 49

Sep 70 43 87

Oct 93 74 122

Nov 95 75 112

Dec 96 76 110

38

Table 4.4 Annual rainfall (mm) Pongola Experiment Station Farm 2046 – 2065

PERCENTILE

YEAR MEAN 10th 90th

2046 74 11 179

2047 86 12 187

2048 74 7 173

2049 69 7 141

2050 60 11 125

2051 53 10 110

2052 69 6 143

2053 59 6 146

2054 67 13 153

2055 66 12 143

2056 83 13 170

2057 67 8 129

2058 71 7 143

2059 58 11 109

2060 69 7 159

2061 65 9 144

2062 64 7 136

2063 68 11 145

2064 61 13 114

2065 69 6 145

39

Figure 4.1 Monthly mean rainfall (mm) Pongola Experiment Station Farm

1967 - 2001

Figure 4.2 Annual mean rainfall

40

Figure 4.3 Monthly mean rainfall

Figure 4.4 Annual mean rainfall

41

The data above are a representation of the annual mean and mean monthly

rainfall in the study area and for the period of interest being the year 1967 – 2062.

This is a micro view of the weather pattern, particularly rainfall, as it plays a major

role in the calculation for of the Weinert N-value.

4.2 DECADAL CHANGE IN EXTREME RAINFALL

The map of southern Africa (Figure 4.5) indicates the different climatic regions.

These are macroclimates and it should be kept in mind that microclimates may

occur within these regions as already discussed in Section 4.1.The map was

adopted from Weinert (1980) by Technical recommendations for Highways.

Currently, other than the Thornthwaite map shown in Figure 2, the Macro Climatic

Regional Map of Southern Africa adopted by TRH4 (1996:40) is the only map used

in the industry in South Africa. It should be noted that TRH4 uses it out of context.

It was derived to determine the weatherability of rocks and the types of clay

minerals formed according to the recorded climate statistics over the up to about

1960 used in the industry. To date, even in light of the imminent threat of climate

change, no credible advances have been made for any improvements on this over

thirty year old design climatic regional map.

42

Figure 4.5: Macro Climatic Regions of southern Africa


In Figure 4.6 all the six different projections indicate an annual increase in extreme

rainfall events in the chosen study area with the exception of the Model for

Interdisciplinary Research On Climate (MIROCmr) portraying a reduction in the

margin of -3 to -4 events per day. The Commonwealth Scientific and Industrial

Research Organisation (CSIRO) mk3.5 ensemble shows the strongest positive

projected change at 3 to 4 events per day. This is at the higher end of the

classification graph and hence confirms the vulnerability of the study area with

regards to storm fall and possible flash flooding.

43

Figure 4.6: Decadal change in extreme rainfall

Source: Piketh et al., (2012)

44

The projected change in the annual frequency of extreme rainfall events shown in

Figure 4.7 is defined as 20 mm of rain falling within 24 hours over an area of

0.5°x0.5° over South Africa, for the period 2071-2100 vs. 1961-1990 (units are

number of events per model grid box per day). The figure shows the ensemble

average of the set of downscaled projections, obtained from six Coupled General

Circulation Model (CGCM) projections of AR4 of the IPCC as reflected by Figure

4.6. This figure could be a possible replacement in future for the current Weinert

N-value.

Figure 4.7: Projected change in the annual frequency of extreme rainfall events

Source: Piketh et al., (2012)

45

4.3 SOIL LABORATORY TESTS AND RESULTS

In this section, the results of tests performed from two soil laboratories (Tshwane

University of Technology and SOILCO) are presented. The first set of results

showed a summary of the six initial samples that were tested (Tables 4.5 to 4.6) at

Tshwane University of Technology. The tabulated results include indicator tests,

maximum dry density and soaked and unsoaked CBR results.

Most of the materials were classified as G10 based on their CBR and PI.

According to Technical Recommendations for Highways (TRH 4) (1996), the

minimum requirement for sub-grade CBR is a soaked CBR of at least 3% at 93%

Mod AASHTO density. The material should also have a maximum swell of 1.5% at

100% Mod AASHTO compaction to ensure that it is not too expansive. This

concurs with the results obtained during the material classification and therefore

the material used for the research was of typical minimum sub-grade quality.

46

Table 4.5 Road P443/1 compacted at varying moisture contents

TEST RESULTS PROJECT No.

ROAD: P443/1

Sample No.

001 002 003 004 Date

02/11/2012 02/11/2012 02/11/2012 02/11/2012

Chainage

0+735 0+960 1+180 0+380 Bound / Lane

L L L L

Section

Bam - Ngwa Bam - Ngwa Bam - Ngwa Bam - Ngwa Layer

SUBGRADE SUBGRADE SUBGRADE SUBGRADE

Soaking method

(96hrs) (96hrs) (96hrs) (96hrs)

SIEVE ANALYSIS (% PASSING) TMH 1: A1(a) & A5 75.000 mm

63.000 mm 53.000 mm

100 37.500 mm

97

26.500 mm

100

91 19.000 mm

95

86

13.200 mm

87 100 59 100 4.750 mm

71 99 47 97

2.000 mm

57 92 36 89 0.425 mm

40 80 23 67

0.075 mm

29 74 15 57

Grading modulus 1.73 0.54 2.26 0.87

ATTERBERG LIMITS Linear shrinkage

2.5 7 3 5

Liquid limit

22 42 24 35 Plastic limit

17 28 18 24

Plasticity index

5 14 6 11

MOD AASHTO Maximum Dry Density (kg/m³) 1958 1965 2018 1713 Optimum moisture content (%) 10.4 10.4 10.6 10.6

CBR

CBR @ 95%

7 1 7 1 CBR @ 93%

5 1 5 1

MATERIAL CLASSIFICATION G10 SPOIL G10 SPOIL CBR @ 90% ; OMC

1 1 2 1

CBR @ 90% ; OMC-4% N/A 14 18 14 CBR @ 90% ; OMC-2% 21 10 8 16 CBR @ 90% ; OMC+2% 2 7 2 11 CBR @ 90% ; OMC+4% 1 7 1 9

47

Table 4.6 Road P435/1 compacted at varying moisture contents


ROAD: P435/1

Sample No.

005 006 Date

03/11/2012 03/11/2012

Chainage Bound / Lane

L L

Section

Skhe - Ndum Skhe-Ndum Layer

B. PIT B. PIT

Soaking method

(96hrs) (96hrs)


63.000 mm

100 100 53.000 mm

94 97

37.500 mm

91 89 26.500 mm

90 86

19.000 mm

85 79 13.200 mm

78 73

4.750 mm

63 58 2.000 mm

57 51

0.425 mm

43 47 0.075 mm

31 43

Grading modulus 1.69 1.59


Liquid limit Plastic limit Plasticity index

NP NP

MOD AASHTO Maximum Dry Density (kg/m³) 1700 1680

Optimum moisture content (%) 9.4 10.6

CBR CBR @ 95%

5 3

CBR @ 93%

4 3 MATERIAL CLASSIFICATION G10 G10 CBR @ 90% ; OMC

2 3

CBR @ 90% ; OMC-4% N/A 15 CBR @ 90% ; OMC-2% N/A 8 CBR @ 90% ; OMC+2% N/A 4 CBR @ 90% ; OMC+4% N/A N/A

48

4.3.1 SOILCO site laboratory results

Another run of tests were performed using a SOILCO Site Laboratory situated in

the vicinity of the two routes P443/1 & P435/1. These tests were performed on two

large samples, each obtained from one of the roads and currently serving as the

pavement sub-grade. Each sample was initially tested for MOD, CBR and

Indicator tests in order to classify it and to determine whether it falls within the

range of subgrade material, being G7 to G10.

Thereafter, each of the samples was quartered into six portions of three moulds

totalling 18 moulds hat would be compacted and soaked in a soaking bath for 0, 2,

3, 5, 6 and 7 day periods (See Appendix D). Swell readings were recorded daily

during soaking period, the specimens were removed from the soaking bath and

then compressed in triplicate as prepared. The results of the second run of tests

are presented in Tables 4.7 to 4.10.

49

Table 4.7 Road P443/1 soaked at varying durations


ROAD: P443/1

Sample No.

657 657 657 657 Date

17/01/2013 17/01/2013 17/01/2013 17/01/2013

Chainage

2+400 2+400 2+400 2+400 Bound / Lane

5.2L 5.2L 5.2L 5.2L

Section

Bam - Ngwa Bam - Ngwa Bam - Ngwa Bam - Ngwa Layer


Soaking period

4 Days 0 Days 2 Days 3 Days


63.000 mm 53.000 mm 37.500 mm 26.500 mm

100 19.000 mm

96

13.200 mm

92 4.750 mm

59

2.000 mm

31 0.425 mm

21

0.075 mm

18

Grading modulus 2.31


5.6

Liquid limit

32 Plastic limit

21

Plasticity index

11


CBR CBR @ 95%

11

CBR @ 93%

17 MATERIAL CLASSIFICATION G7 G7 G7 G7

CBR @ 90% ; OMC

24 21 20

50



ROAD: P443/1

Sample No.

657 657 657 Date

17/01/2013 17/01/2013 17/01/2013

Chainage

2+400 2+400 2+400 Bound / Lane

5.2L 5.2L 5.2L

Section

Bam - Ngwa Bam - Ngwa Bam - Ngwa Layer

SUBGRADE SUBGRADE SUBGRADE

Soaking period

5 Days 6 Days 7 Days


63.000 mm 53.000 mm 37.500 mm 26.500 mm 19.000 mm 13.200 mm 4.750 mm 2.000 mm 0.425 mm 0.075 mm

Grading modulus



MOD AASHTO Maximum Dry Density (kg/m³) 1950 1950 1950

Optimum moisture content (%) 8.6 8.6 8.6

CBR CBR @ 95%

CBR @ 93% MATERIAL CLASSIFICATION G7 G7 G7

CBR @ 90% ; OMC

18 17 13

51



ROAD: P435/1

Sample No.

658 658 658 658 Date

17/01/2013 17/01/2013 17/01/2013 17/01/2013

Chainage

8+000 8+000 8+000 8+000 Bound / Lane

3.0R 3.0R 3.0R 3.0R

Section

Skhe - Ndumo Skhe - Ndumo Skhe - Ndumo Skhe - Ndumo Layer


Soaking period

4 Days 0 Days 2 Days 3 Days


63.000 mm 53.000 mm 37.500 mm 26.500 mm

100 19.000 mm

99

13.200 mm

98 4.750 mm

97

2.000 mm

94 0.425 mm

80

0.075 mm

70

Grading modulus 0.56 3.00


3.7

Liquid limit

28 Plastic limit

21

Plasticity index

7


CBR CBR @ 95%

CBR @ 93% MATERIAL

CLASSIFICATION G7 G7 G7 G7 CBR @ 90% ; OMC

22 21 19

52



ROAD: P435/1

Sample No.

658 658 658 Date

17/01/2013 17/01/2013 17/01/2013

Chainage

8+000 8+000 8+000 Bound / Lane

3.0R 3.0R 3.0R

Section

Skhe - Ndumo Skhe – Ndumo Skhe - Ndumo Layer

SUBGRADE SUBGRADE SUBGRADE

Soaking period

5 Days 6 Days 7 Days


63.000 mm 53.000 mm 37.500 mm 26.500 mm 19.000 mm 13.200 mm 4.750 mm 2.000 mm 0.425 mm 0.075 Mm

Grading modulus



MOD AASHTO Maximum Dry Density (kg/m³) 1889 1889 1889

Optimum moisture content (%) 8.2 8.2 8.2

CBR CBR @ 95%

CBR @ 93% MATERIAL CLASSIFICATION G7 G7 G7

CBR @ 90% ; OMC

17 13 12

53

4.3.2 Graphical presentation of SOILCO site laboratory CBRs

The results of the SOILCO site laboratory CBR’s were plotted against the number

of days that each specimen was soaked in a soaking bath. This was done for both

samples taken from P443/1 and P435/1. Various trendlines were then calculated

to determine the best-fit model for the data. A linear trendline on both graphs

provided similar models with R2 values of 0.945 and 0.985 for route P443/1 and

P435/1 respectively. The best-fit models are (Sample No. 657);

886.25*9563.1 dCBRd Equation 1

Sample No. 658

967.24*1718.2 dCBRd Equation 2

where

CBRd = CBR at day number d d = day for which CBR is predicted

Tables 4.11 and 4.12 show the CBR results that are plotted on Figures 4.8 and

4.9. Results of the CBR value predicted using models are also shown in the

tables. It should also be noted that when the wet days as obtained from Figure 4.8

and 4.9 are used, the formulae can give out the predicted subgrade strength for

the said number of wet days.

54

Table 4.11 Road P443/1actual and predicted CBR values

Road P443/1

Time (Days)

CBR (%)

Predicted CBR (%)

0 24 25.9 2 21 22.0 3 20 20.0 5 18 16.1 6 17 14.1 7 13 12.2

Figure 4.8 Variation of CBR with time of soaking

55

Table 4.12 Road P435/1 actual and predicted CBR values

Road P435/1

Time (Days)

CBR (%)

Predicted CBR (%)

0 22 25.0 2 21 20.6 3 19 18.5 5 17 14.1 6 13 11.9 7 12 9.8

Figure 4.9 Variation of CBR with time of soaking

56

4.4 PROJECTED CLIMATE CHANGE

Rainfall simulations from the CSAG and the CSIR presented in chapter 4 indicate

a similar pattern to current conditions with an almost negligible decrease. The

predicted general reduction in rainfall is about 0.2% compared with present day

scenarios. The winter season (May to August), however, is projected to have a

sharper decrease in rainfall of about 5% which translates to around 80 mm less in

the 4 months.

Judging from the data presented in the previous chapters, a projection of 50 years

into the future signals minimal climate change in terms of rainfall; however, natural

weather variability threatens to be the dominant signal. The number of wet days

will most likely have more bearing on future performance of the subgrade than the

actual rainfall in millimetres.

4.5 EFFECT OF EXTREME RAINFALL ON SOUTHERN AFRICAN REGION

A review of historical records was also carried out to get a general idea of the

variability of rainfall with special reference to the extremes. The extremes are

important because recent history warns of parameters such as rainfall frequency,

intensity and duration to change much more quickly than the mean.

Over the whole of South Africa, increases in the magnitude of extreme daily

rainfall events with return periods of 10 and 30 years have been projected.

Noteworthy is the increase of these events even in regions where reductions in the

mean annual rainfall were forecasted.

57

In general, East Africa is projected to become generally wetter, whilst southern

Africa is projected to become generally drier with a relatively strong signal of

drying projected for Zimbabwe, Zambia and Angola. The Highveld and central

interior of South Africa is projected to become somewhat wetter, despite the

general drying signal projected for southern Africa.

According to Piketh et al., (2012), projected changes in extreme rainfall events in

South Africa include wet-spells and widespread flooding over the South African

Highveld and this may result from a number of different weather systems (or from

a combination of different weather systems). The Highveld constitutes parts of the

Mpumalanga, Northern Cape, North West, and Limpopo provinces, and virtually all

of Gauteng and the northern Free State. It covers an area of almost 400,000 km²,

or roughly thirty percent of South Africa's land area. Such extreme rainfall events

have already been occurring for periods over a decade now in Mozambique and

the Mpumalanga province.

4.6 EFFECT OF MOISTURE ON CBR TEST

As soils in construction can only be compacted properly at OMC, the initial method

of using variable compaction moisture contents during compaction of the samples

did not satisfy the intended purpose of the research. Therefore another round of

CBR tests was conducted; focusing on different soaking periods for samples

compacted at OMC. It also focused on soaking of the CBR samples done at varying

durations in order for this failure relationship to be developed between the soaked

and unsoaked samples. This is also due to the fact that the CBR test is the only test

http://en.wikipedia.org/wiki/Mpumalanga

http://en.wikipedia.org/wiki/Northern_Cape

http://en.wikipedia.org/wiki/North_West_(South_African_province)

http://en.wikipedia.org/wiki/Limpopo

http://en.wikipedia.org/wiki/Gauteng

http://en.wikipedia.org/wiki/Free_State_(South_African_province)

58

that can give worst case scenario strength of the soil material, hence the use of

such a test in this study.

According to Emery (2001), there are two methods to estimate how material will

perform when it is soaked; viz

i. take a sample, and do a laboratory soaked CBR test

ii. check the Plasticity Index (PI; from the Atterberg Limit tests). Low PI

materials (PI < 6) will not weaken too much when wet but higher PI materials

will weaken significantly when wet.

Both of these tests have been performed on all of the soil samples taken with

inconclusive results on the initial test run but more conclusive results on the

second battery of tests. The PI for the road samples was in the range of 7-11.

Initially, due to improper test procedures, six samples were compacted at varying

moisture contents (OMC, OMC+2%, OMC+4%, OMC-2% and OMC-4%).

This yielded very low CBR results during penetration of the specimens as they

were compacted below and above OMC. For the specimens compacted above

OMC, their low strengths were attributed to the increased moisture content during

both the compaction and subsequently compression processes as increased

moisture reduces the particle interlock. The specimens with less than OMC during

compaction were considered to have had very high voids content which promoted

an excessive amount of water to fill the voids and also resulting in reduced

strength during compression. It is for these moisture variation effects therefore that

specimens for CBR determination are compacted at OMC.

59

4.7 INTERPRETATION OF STRENGTH AND FORMULAE USAGE

The second run of tests required all of the samples to be compacted at optimum

moisture content but soaked for varying periods of time. The results were then

tabulated and plotted on a chart with the CBR values against the soaking period. A

linear trend was established for each of the samples’ soaked CBR values and a

formula representing the data was deduced. These formulae indicated that

unsoaked specimen can be used to obtain the four day soaked laboratory CBR

without the soaking process for these or similar materials. It should, however, be

borne in mind that there would be no swell measurements as the specimens would

be unsoaked. As seen in Figures 4.8 and 4.9 and judging by the similarity of the

intercepts and gradients in equations 1 and 2, a new equation representing both

materials can also be derived as follows.

25*2 dCBRd Equation 3

where;

d = day for which CBR is predicted

The derivation of Equation 3 is aimed at simplifying Equations 1 and 2; thereby

enabling the CBR strength of materials to be obtained within a shorter time frame

than the current five day period it takes to compact, soak and penetrate the soil

specimen. An example of the predicted values based on equation 3 is shown in

Tables 4.13 and 4.14. The value of R2 as shown on Figure 4.8 and Figure 4.9 is

0.9448 and 0.985 for samples 657 and 658 respectively. The closeness of R2 to

60

1.0 confirms a good fit to the results of CBR tests obtained under different times of

soaking.

Table 4.13: Road P443/1 actual and predicted CBR values (composite

equation)

Road P443/1 Time

(Days) CBR (%)

Predicted CBR

0 24 25.0 2 21 21.0 3 20 19.0 5 18 15.0 6 17 13.0 7 13 11.0

Table 4.14: Road P435/1 actual and predicted CBR values (composite

equation)

Road P435/1 Time

(Days) CBR (%)

Predicted CBR

0 22 25.0 2 21 21.0 3 19 19.0 5 17 15.0 6 13 13.0 7 12 11.0

The equation is ideal for use on weaker gravels used as subgrade for road

pavements. These are materials that generally have a CBR strength ranging from

3% to 15% and often have a PI greater than 12%. The formula may not always be

useful or accurate, especially when dealing with other gravels considered stronger

than typical subgrade materials. The equation, similarly to other test methods such

as DCP, provides a rapid and accurate test to estimate the CBR of subgrades. It is

seen as ideal to expedite the four day testing period particularly where results are

required speedily.

61

4.8 RESULTS WITH REFERENCE TO OBJECT OF INVESTIGATION

The intention of this section is to produce findings which, as the title suggests, are

to assess the effects of climate change on road subgrades. This has been

accomplished in the investigation by extended soaking of the subgrade specimens

from initial compaction at optimum conditions. Thereafter, performing the

penetrations on the samples soaked for a varying number of days to record the

change in CBR value as time progresses.

Ideally this should be done on new road construction where there is strict control

on the compaction of the subgrade to 90% MOD AASHTO density, where a CBR

test under un-soaked conditions will give a maximum strength. This research was

carried out without the luxury of new road construction; as such the investigation

has been concentrated on the rehabilitation of an existing pavement involving the

in-situ stabilisation of the existing road surfacing and base. The upgraded

pavement is shown in Table 4.15.

Table 4.15: P443 and P435/1 pavement design

LAYER DESCRIPTION

Surfacing 40mm Continuously graded asphalt.

Base 200mm

Gravel CBR>25% at 95% Stabilised (C3) compacted to 97% of modified AASHTO density, UCS: 1.5 to 3.0 MPa at 100% Mod AASHTO, minimum ITS: 250Kpa, Maximum PI: 6% after stabilisation.

Sub base In-situ Nat. Gravel CBR>19% at 95% (G6) compacted to 95% of modified AASHTO density, PI < 10 or 3GM +10, Maximum swell 1% @ 100% Mod AASHTO.

62

In-situ density tests carried out on the subgrade layer where the CBR tests have

been undertaken have yielded a relative compaction of 91.4% for Road P435,

which does not differ greatly from the assumed 90% MOD AASHTO density. On

Road P443, however, an average relative compaction of 94.3% MOD AASHTO

density was recorded. This would indicate a considerable increase in strength

however relative compaction values from seven test pits along the road carried out

as part of the centreline investigation range from 80.5% to 99.3% MOD AASHTO

density. For the purpose of this report the in-situ relative compaction is assumed to

be 90% relative compaction in both cases. On completion of the findings a

separate review of the situation at the test pit position on Road P435 is given for

comparison.

4.9 ASSUMED ELASTIC MODULI FOR VARIATION IN CBR AND MOISTURE

CONTENT

As the in-situ moisture contents increase, there is, as shown, a corresponding

reduction in CBR values. The CBR test gives an indirect measure of shear

strength, which in turn can be used to estimate the elastic moduli of granular

materials as derived for the pavement design catalogue, and form the basis of

Draft TRH 4 (1996). A linear model has been developed between the increase in

moisture content and the reduction in CBR at 90% MOD AASHTO density. This

implies a reduction in strength, which is measured by the elastic moduli of the

materials. From the moduli values used in TRH 4 (1996), the following moduli

values in relation to the CBR values at 90% MAASHTO density have been

structured (Table 4.16).

63

Using the South African design procedures, the variation in pavement carrying

capacity is estimated and it shows a reduction in CBR for prolonged moisture

conditions and subsequent reduction in predicted life.

Table 4.16: Moduli Values versus CBR Values at 90% MAASHTO density

Source: Theyse (1995:3)

4.10 REQUIRED DESIGN LIFE

The data obtained is summarised in Table 4.17

Table 4.17: Traffic Data Available

Station No Location AADT % H V’s Date

2489A West of P522-2 1282 8.3 Jan 2007

CBR @ 90% MAASHTO E (MPa)

25

20

15

10

7

5

3

200

165

130

95

75

60

45

64

The projected design traffic for route P443 and P435 is as follows;

15 years 0.85X106 E80’s

20 years 1.18X106 E80’s

It is intended that a 20 year design life be adopted. Due to the higher CBR values

at 90% compaction the initial design life will be much higher, but reduce

considerably with increase in moisture.

4.11 SIMULATED PAVEMENT CARRYING CAPACITY IN RELATION TO SUB-

GRADE MOISTURE

With reference to the data provided in Tables 4.11 and 4.12 and Figures 4.8 and

4.9 respectively the predicted pavement carrying capacity was determined by

equation 4 as obtained from TRH 4 (1996) reported in Tables 4.18 and 4.19 Time

for 0 days refers to optimum conditions for CBR penetration on a “dry” sample.

)(/'80% yfxHVsExHVxTrafficpacityPavementCa Equation 4

where;

Traffic = vehicles per day per lane

%HV = percentage of heavy vehicles per day

E80’s/HV = E80 per heavy vehicle

fy = traffic growth factor

65


content; P443

Time (Days) CBR @ 90% Carrying capacity(106 E80’s)

0

2

3

4

5

6

7

10

25.9

22

20

18

16

14

12

6

2.73

2.56

2.43

2.28

2.19

1.98

1.90

1.38

It can be seen that after a fall in the CBR to around 23% of its original value the

pavement carrying capacity has fallen to almost half of its original value, but still

achieves a 20 year life requirement. Had it been analysed on its measured in-situ

strength value the life would have fallen from 3.43X106 E80’s to 2.73X106 E80’s,

which would never have posed a threat to the required carrying capacity.

66


content for P435

Time (Days) CBR @ 90% Carrying capacity (106 E80’s)

0

2

3

4

5

6

7

10

25

21

19

18

17

13

12

3

2.73

2.49

2.35

2.28

2.23

1.94

1.90

0.87

After 10 days of continual soaking the road would only have provided for a 15 year

life as the carrying capacity is reduced to a third of its original capacity.

67

CHAPTER 5: CONCLUSION AND RECOMMENDATIONS

5.1 CONCLUSION

5.1.1 Macro climatic regions of southern Africa

The results produced by the atmospheric global circulation model CCAM of the

Commonwealth Scientific and Industrial Research Organisation (CSIRO), in

Australia through (Piketh et al., 2012) is a true graphical representation of the

projected change in extreme rainfall for the next 86 years. In light of the presented

results, it is clear that adoption of these findings is likely to become imperative for

the development and precision of engineering designs. This is partly due to the

fact that the current map, adopted from Weinert, (1974) is an old map that can no

longer be truly relied on in this changing world climate. Furthermore, road

pavements are ordinarily designed for timespans of over two decades; therefore,

the continued use of such a dated map poses a threat to the integrity of these

pavements which are designed for future generations. This approach was

established by understanding the role of climate variability especially with regards

to extreme rainfall events.

The new map as shown in Figure 4.7 will be a very helpful starting point for the

precise determination of future Weinert N-Values. It is important to note that the

resultant map indicates the extreme rainfall patterns and assumes that the

potential evaporation is constant throughout the entire region. The Weinert N-

Value can be calculated at any point where climatic data is available in the

southern African region, even with the current old map as attested by TRH 4

68

(1996). The difference however is that the map presented in this study is based on

future simulation of storm intensity and not ordinary rainfall.

The ratio of evaporation to rainfall which gives us the N-Value as described by

Weinert (1974) is clearly no longer properly represented from the then map which

had a vast majority of the southern African land classified as dry. In future, there

are likely to be more moderate and wet areas than when the map was developed.

Areas characterised as dry, such as the Western Cape and the Karoo, shall be

now described as moderate. Another visible change on the map is the increase

from three different climatic regions to six. Of these six climatic regions, some are

considered borderline and may be interpreted as a grey area by many engineers.

5.1.2 Rapid assessment of CBR

The equations developed will serve as a way to obtain the CBR of a material

within a shorter timeframe coupled with the added advantage of utilising the same

apparatus as the conventional laboratory tested samples. It will be advisable for

any practising engineer on site to utilise this method as it is a quick and relatively

easy method of predicting a preliminary CBR whilst awaiting the official CBR result

which requires at least 5 days. This minimum of 5 days waiting period can also be

expressed as time lost on site during the normal testing process. Furthermore, this

method takes place concurrently with the conventional CBR method thus reducing

any delays that would have resulted from a completely new method. The cost of

developing new apparatus, calibration procedures and manufacturing which would

69

have been coupled with this rapid test has also been eliminated through the use of

already existing procedures/apparatus.

5.1.3 Reduced pavement carrying capacity

The South African design procedures used for the calculation of pavement

carrying capacity variations proved a trend similar to the reduction in CBR for

prolonged moisture conditions. It can be seen from Table 4.18 that after the CBR

falls to around 23% of its original value; the pavement carrying capacity of road

P443 has fallen to almost half of its original value, but still achieves a 20 year life

requirement. This implies that extended soaking definitely has a greater negative

effect on the strength of the material than would normally have been expected.

However, upon analysis of its measured in-situ strength value, the carrying

capacity of the pavement falls from 3.43X106 E80’s to 2.73X106 E80’s, which

would not pose a threat to the required design life.

According to the E80’s described above, the pavement category will, however,

change from a major interurban freeway (ES10) to an interurban collector or rural

road (ES3) whereby the design reliability changes from 95% to 90% due to the

extended soaking which lasted 7 days. This proves beyond doubt that climate

change in terms of storm intensity as demonstrated in the research through

extended soaking of specimens will definitely have a negative effect on inter alia

the design life, pavement category and design reliability.

70

5.2 RECOMMENDATIONS

A number of suggestions for further study are as follows.

i. Utilization of the actual evaporation rates at all southern African weather

stations for the calculation of the Weinert N-Value. These must be obtained

from a reputable source such as the CSIR, the Climate Systems Analysis

Group (CSAG) or the Agricultural Research Council (ARC).

ii. Plotting these calculated N-Values on a Southern African map

iii. Re-calculation of the N-Values using the six new climatic regions

Although these findings are based on subgrade soil, a suggestion for further study

is to extend this methodology to a wide range of other materials that are not only

of subgrade quality in an attempt to achieve the goal of a shorter timeframe for all

laboratory soaked CBRs and hopefully materials classification. It is current

practice that most testing of subgrade materials during design uses the soaked

California Bearing Ratio (CBR) for paved roads and it is recommended that un-

soaked (field) CBR values should be used particularly in dry regions. This

recommendation is strongly contradicted by the findings of this study as it requires

no extended soaking for both wet and dry regions. Furthermore, given the proven

climate change and subsequently the change in climatic regions, it therefore

warrants that current regions be verified against this study’s newly established

climatic regional map prior to finalising all design and environmental influences.

This agrees with the recent strong move by road designers towards

environmentally optimised pavement design using expected in situ moisture

71

contents which relies on a good and well-maintained drainage system but that

save large amounts of money.

This study also established that the formula can be used with confidence to a

certain number of days as accuracy cannot be guaranteed when time of soaking

exceeds seven days which was the maximum period of soaking used in this

research. It follows that the use of the formula should be considered not to be

absolute, but rather comparative with the orthodox four day CBR method; and

should be used with care.

Finally, additional change in moisture should be investigated on various material

types as there has been suspicion for some time by designers that the four day

soaking does not allow adequate penetration of water through all materials,

particularly high density materials.

72

REFERENCES

Bureau for Industrial Co-operation, 2011. Chapter 3 - Harmonisation of Pavement

Design Standards [Online] Available from:

http://www.eac.int/infrastructure/index.php?option=com_docman&Itemid=158

[Accessed: 11/04/2014]

CARRERA, A., DAWSON, A. and STEGER, J., 2010. State of the art of materials’

sensitivity to moisture change [Online] Available from:

http://www.eranetroad.org/index.php?option=com_docman&task=cat_view&gid=8

8&Itemid=53 [Accessed: 12/07/2012]

CARTWRIGHT, A., 2008. Phase three: Final Report. A Sea-Level Rise Risk

Assessment for the City of Cape Town [Online] Available from:

http://www.capetown.gov.za/en/EnvironmentalResourceManagement/publications/

Documents/Phase%203%20-%20A%20Sea-

Level%20Rise%20Risk%20Assessment%20(SLRRA).pdf [Accessed: 20/03/2014]

CHINOWSKY, P., SCHWEIKERT, A., STRZEPEK, N., MANAHAN, K.,

STRZEPEK, K., and SCHLOSSER, C.A. 2011. Adaptation Advantage to Climate

Change Impacts on Road Infrastructure in Africa through 2100 [Online]. Available

from: http://www.wider.unu.edu/.../working-papers/2011/en.../wp2011-025.pdf

[Accessed: 17/02/2013]

http://www.eac.int/infrastructure/index.php?option=com_docman&Itemid=158

http://www.eranetroad.org/index.php?option=com_docman&task=cat_view&gid=88&Itemid=53

http://www.eranetroad.org/index.php?option=com_docman&task=cat_view&gid=88&Itemid=53

http://www.capetown.gov.za/en/EnvironmentalResourceManagement/publications/Documents/Phase%203%20-%20A%20Sea-Level%20Rise%20Risk%20Assessment%20(SLRRA).pdf



http://www.wider.unu.edu/.../working-papers/2011/en.../wp2011-025.pdf

73

Climate Institute, 2012. Oceans and Sea Level Rise - Consequences of Climate

Change on the Oceans [Online] Available from: http://www.climate.org/topics/sea-

level/index.html [Accessed: 18/02/2013]

Climatology and Climate Change, 2009 [Online]. Available from:

http://www.joburg-

archive.co.za/2009/pdfs/report_evironment/enviro_climatology.pdf. [Accessed:

16/02/2013]

DAVIES, B. 2004. A model for the prediction of sub-grade soil resilient modulus for

flexible pavement design: Influence of moisture content and climate change. M.Sc.

dissertation, University of Toledo.

Department of Environmental Affairs and Tourism. 2005. Global Climate Change

and Ozone Layer Protection – What does it mean for South Africa [Online].

Available from:

http://www.environment.gov.za/climatechange2005/What_does_it_mean_for_Sout

h_Africa.htm [Accessed: 05/01/2013].

DIGITAL MAP STUDIO. 2014. [Online] Available from:

http://www.customdigitalmaps.com/free-maps.htm [Accessed: 18/09/2014]

EMERY, S. 2001. DCP Testing and analysis [Online] Available from: http://www.geocities.com/profemery.info/pavement/DCP.doc [Accessed 23/02/2013]

http://www.climate.org/topics/sea-level/index.html

http://www.climate.org/topics/sea-level/index.html

http://www.joburg-archive.co.za/2009/pdfs/report_evironment/enviro_climatology.pdf

http://www.joburg-archive.co.za/2009/pdfs/report_evironment/enviro_climatology.pdf

http://www.environment.gov.za/climatechange2005/What_does_it_mean_for_South_Africa.htm

http://www.environment.gov.za/climatechange2005/What_does_it_mean_for_South_Africa.htm

http://www.customdigitalmaps.com/free-maps.htm

http://www.geocities.com/profemery.info/pavement/DCP.doc

74

FAIRHURST, L. 2008. Global Climate Change and Adaptation– A Sea-Level Rise

Risk Assessment. [Online]. Available from:

http://www.capetown.gov.za/en/EnvironmentalResourceManagement/publications/

Pages/Reportsand.aspx [Accessed: 20/02/2012].

Guidelines for Human Settlement Planning and Design [Online] Available from:

http://www.csir.co.za/Built_environment/RedBook/ [Accessed: 20/02/2012]

Impact of climate change on road infrastructure-Austroads, 2004 [Online].

Available from: http://www.btre.gov.au/info.aspx?ResourceId=692andNodeId=136

[Accessed: 15/02/2013].

KANNEMEYER, L. 2010. State of South Africa’s Road Network [Online] Available

from: http://www.sarf.org.za/.../2_1115-1145%20%20L%20Kannemeyer.pdf

[Accessed: 15/02/2014]

KOCH, W. 2011. Fastest sea-level rise in 2,100 years linked to climate change.

USA Today, Jun. 21

LI, Q., MILLS, L. and MCNEIL, S. 2011. The Implications of Climate Change on

Pavement Performance and Design [Online] Available from:

http://www.ce.udel.edu/UTC/20110926_FinalReport_Pavement_ClimateChange.p

df [Accessed: 18/02/2012]

http://www.capetown.gov.za/en/EnvironmentalResourceManagement/publications/Pages/Reportsand.aspx

http://www.capetown.gov.za/en/EnvironmentalResourceManagement/publications/Pages/Reportsand.aspx

http://www.csir.co.za/Built_environment/RedBook/

http://www.btre.gov.au/info.aspx?ResourceId=692andNodeId=136

http://www.sarf.org.za/.../2_1115-1145%20%20L%20Kannemeyer.pdf

http://www.ce.udel.edu/UTC/20110926_FinalReport_Pavement_ClimateChange.pdf

http://www.ce.udel.edu/UTC/20110926_FinalReport_Pavement_ClimateChange.pdf

75

MENDEL, G. 2006. Climate change, urban flooding and the rights of the urban

poor in Africa [Online]. Available from:

http://www.tiempocyberclimate.org/portal/archive/pdf/tiempo64low.pdf [Accessed:

16/02/2013].

MNDAWE, M. B. NDAMBUKI, J. M. and KUPOLATI, W. K, 2013. Revision of the

macro climatic regions of southern Africa. In: Proceedings of the 6th Africa

transportation technology Transfer (T2) Conference, March 4-8, 2013, Gaborone,

Botswana

NOREM, H. and MÖLLER, S. 2007. Climate Change and Road Management

[Online] Available from: http://www.nordicroads.com/wp-

content/uploads/2012/10/2-2007.pdf [Accessed: 15/02/2012].

Pavement Age [Online]. Available from:

http://www.nra.co.za/live/content.php?Item_ID=65 [Accessed: 15/02/2013].

PIKETH, S., FATTI, C., AKOON, I., BURGER, R., DUNSMORE, S.,

ENGELBRECHT F., SWIGGERS, C., and van WYK, F. 2012. The Impact of

Climate Change on Water Service Delivery. (Paper). Unpublished

Revision of the South African Flexible Pavement Design Method: Mechanistic-

empirical method [Online] Available from:

http://researchspace.csir.co.za/dspace/handle/10204/1323 [Accessed:

10/07/2012]

http://www.tiempocyberclimate.org/portal/archive/pdf/tiempo64low.pdf

http://www.nordicroads.com/wp-content/uploads/2012/10/2-2007.pdf

http://www.nordicroads.com/wp-content/uploads/2012/10/2-2007.pdf

http://www.nra.co.za/live/content.php?Item_ID=65

http://researchspace.csir.co.za/dspace/handle/10204/1323

76

ROLT, J., ACQUAH, B., and DONE, S., 2004. Environmentally optimised design:

analysis of road performance in Ghana S.l.: s.n.

Sabita. 2011. vol. 25. 2nd Issue [Online]. Available from:

http://www.sabita.co.za/documents/asnews2-11.pdf [Accessed 24/04/2014]

SAICE Infrastructure Report Card on South Africa 2011 [Online]. Available from:

http://www.csir.co.za/enews/2011_jun/download/infrastructure_report_card_sa_20

11.pdf [Accessed: 17/02/2012]

SCHAEFER, V.R. 2008. Design Guide for Subgrades and Subbases [Online]

Available from: http://www.ctre.iastate.edu/research/reports.cfm [Accessed:

12/04/2012].

State of the environment report. S.a . [Online]. Available from:

http://www.ngo.grida.no/soesa/nsoer/issues/coast/state.htm [Accessed:

15/02/2012].

Technical Recommendations for Highways. 1996. Technical Recommendations

for Highways 4, Department of Transport, Pretoria: Government Printer.

Technical Recommendations for Highways. 1985. Technical Recommendations

for Highways 14, Department of Transport, Pretoria: Government Printer.

http://www.sabita.co.za/documents/asnews2-11.pdf

http://www.csir.co.za/enews/2011_jun/download/infrastructure_report_card_sa_2011.pdf

http://www.csir.co.za/enews/2011_jun/download/infrastructure_report_card_sa_2011.pdf

http://www.ctre.iastate.edu/research/reports.cfm

http://www.ngo.grida.no/soesa/nsoer/issues/coast/state.htm

77

The World Bank. S.a [Online] Available from:

http://www.worldbank.org/transport/roads/conandmain.htm [Accessed:

18/02/2012]

THEYSE, H.L. 1995. Technical Recommendations for Highways 4 Revision Phase

2: Mechanistic Design Analysis of the pavement structures contained in TRH 4

(1995); Pavement Design Catalogue. Department of Transport: Pretoria:

Government Printer.

THEYSE, H.L, MAINA, J.W. and KANNEMEYER, L. 2007. Revision of the South

African flexible pavement design method: proceedings of the 9th Conference on

Asphalt Pavements for Southern Africa, held in Botswana on 2 -5 September,

2007. Gaborone: Document Transformation Technologies cc

Trademark Southern Africa. 2011. Climate change impacts on road infrastructure

in Africa through 2100 [Online]. Available from:

http://www.trademarksa.org/news/climate-change-impacts-road-infrastructure-

africa-through-2100 [Accessed: 17/02/2012]

TRL Limited. S.a. Rational road drainage [Online]. Available from:

http://www.transport-links.org/transport_links/filearea/documentstore/119_PR-INT-

244.PDF [Accessed 25/06/2014]

http://www.worldbank.org/transport/roads/con&main.htm

http://www.trademarksa.org/news/climate-change-impacts-road-infrastructure-africa-through-2100

http://www.trademarksa.org/news/climate-change-impacts-road-infrastructure-africa-through-2100

http://www.transport-links.org/transport_links/filearea/documentstore/119_PR-INT-244.PDF

http://www.transport-links.org/transport_links/filearea/documentstore/119_PR-INT-244.PDF

78

Washington Asphalt Pavement Association [Online] Available from:

http://www.asphaltwa.com/2010/09/17/design-factors-environment/ [Accessed:

18/02/2012]

WEINERT, H.1980. Natural road construction materials of southern Africa. CSIR.

YOUMAN, P. 2007. The Implications Of Climate Change On Road Infrastructure

Planning, Design And Management [Online] Available from:

http://www.coastalconference.com/2007/papers2007/Paul%20Youman.pdf

[Accessed: 18/02/2012]

http://www.asphaltwa.com/2010/09/17/design-factors-environment/

http://www.coastalconference.com/2007/papers2007/Paul%20Youman.pdf

79

APPENDICES

80

Appendix A: Mount Kilimanjaro Ice Cap (1993 vs. 2000)

Source: World Culture Pictorial (2013)

81

Appendix B: Flood damage to Coleraine Drive, Sandton

Source: Sunday Times (2014)

82

Appendix C: Road P443/1

83

Appendix D: Sample quartering by cone method

84

Appendix E: Sample compaction by mechanical means

85

Appendix F: Soaking of specimen in a soaking bath

86

Appendix G: Specimen compression for CBR determination