Issue 12, 2013 Journal of Civil Engineering and Architecture

Journal of Civil Engineering
and Architecture
Volume 7, Number 1 , December 2013 (Serial Number 73) 2
David Publishing Company
www.davidpublishing.com
Publishing David

Publication Information:
Journal of Civil Engineering and Architecture is published monthly in hard copy (ISSN 1934-7359) and online (ISSN
1934-7367) by David Publishing Company located at 3592 Rosemead Blvd #220, Rosemead, CA 91770, USA.

Aims and Scope:
Journal of Civil Engineering and Architecture, a monthly professional academic journal, covers all sorts of researches
on structure engineering, geotechnical engineering, underground engineering, engineering management, etc. as well
as other issues.

Editorial Board Members:
Dr. Tamer A. El Maaddawy (Canada), Prof. San-Shyan Lin (China Taiwan), Dr. Songbai Cai (China), Dr. Xiaoyan Lei
(China), Prof. Vladimir Patrcevic (Croatia), Dr. Sherif Ahmed Ali Sheta (Egypt), Prof. Nasamat Abdel Kader (Egypt),
Prof. Mohamed Al-Gharieb Sakr (Egypt), Prof. Olga Popovic Larsen (Denmark), Prof. George C. Manos (Greece), Dr.
Konstantinos Giannakos (Greece), Pakwai Chan (Hong Kong), Dr. K. Muthukkumaran (India), Chiara Vernizzi (Italy),
Prof. Michele Maugeri (Italy), Dr. Giovanna Vessia (Italy), Prof. Valentina Zileska-Pancovska (Macedonia), Dr. J.
Jayaprakash (Malaysia), Mr. Fathollah Sajedi (Malaysia), Prof. Nathaniel Anny Aniekwu (Nigeria), Dr. Marta Sowik
(Poland), Dr. Rafael Aguilar (Portugal), Dr. Moataz A. S. Badawi (Saudi Arabia), Prof. David Chua Kim Huat
(Singapore), Dr. Vail Karakale (Waiel Mowrtage) (Turky), Dr. A.Senem Deviren (Turkey), Dr. Yasemin Afacan
(Turkey), Dr. Ming An (UK), Prof. Ahmed Elseragy (UK), Prof. Jamal Khatib (UK), Dr. John Kinuthia (UK), Dr.
Johnnie Ben-Edigbe (UK), Dr. Yail Jimmy Kim (USA), Dr. Muang Seniwongse (USA), Prof. Xiaoduan Sun (USA), Dr.
Zihan Yan (USA), Dr. Tadeh Zirakian (USA).

Manuscripts can be submitted via Web Submission, or E-mail to civil@davidpublishing.com or
civil@davidpublishing.org. Submission guidelines and Web Submission system are available at
http://www.davidpublishing.com, www.davidpublishing.org.

Editorial Office:
3592 Rosemead Blvd #220, Rosemead, CA 91770, USA
Tel: 1-323-984-7526, 323-410-1082 Fax: 1-323-984-7374, 323-908-0457
E-mail: civil@davidpublishing.com; civil@davidpublishing.org; shelly@davidpublishing.com.

Copyright2013 by David Publishing Company and individual contributors. All rights reserved. David Publishing
Company holds the exclusive copyright of all the contents of this journal. In accordance with the international
convention, no part of this journal may be reproduced or transmitted by any media or publishing organs (including
various websites) without the written permission of the copyright holder. Otherwise, any conduct would be
considered as the violation of the copyright. The contents of this journal are available for any citation. However, all
the citations should be clearly indicated with the title of this journal, serial number and the name of the author.

Abstracted / Indexed in:
Database of EBSCO, Massachusetts, USA
Chinese Database of CEPS, Airiti Inc. & OCLC
Cambridge Science Abstracts (CSA)
Ulrichs Periodicals Directory
Summon Serials Solutions
Google Scholar
ProQuest

Subscription Information:
$520/year (print) $360/year (online) $680/year (print and online)

3592 Rosemead Blvd #220, Rosemead, CA 91770, USA
Tel: 1-323-984-7526, 323-410-1082 Fax: 1-323-984-7374, 323-908-0457
E-mail: civil@davidpublishing.com; civil@davidpublishing.org; shelly@davidpublishing.com.

www.davidpublishing.com
DAVID PUBLISHING
D

Jour nal of Ci vi l Engi neeri ng
and Archi t ect ure

Volume 7, Number 12, December 2013 (Serial Number 73)

Contents
Engineering Structures
1471 Consideration of Shape Characteristics for Apartments Incorporating Daylight and Energy
Performance
Ji-Eun Lee and Kang Up Lee
1485 Performance Evaluation of New Two Axes Tracking PV-Thermal Concentrator
Luis C. Chea, Hkan Hkansson and Bjrn Karlsson
1494 A Study on the Architectural Application of Aerogel
Young Cheol Kwon
Structural Engineering
1501 Development of Design Response Spectra Based on Various Attenuation Relationships at Specific
Location
Hapsa Husen, Taksiah Abdul Majid, Fadzli Mohamed Nazri, Mohd Rashwan Arshad, Ade Faisal and
Siti Khadijah Che Osmi
1507 Structural Retrofitting of Old Illyria Hotel
Feti Selmani and Musa Stavileci
1514 Challenges of Structure in Todays Architectural, Economic and Social Context
Jorge Bernabeu Larena and Alejandro Bernabeu Larena
1522 Simulation of Corrosion Crack in Pretensioned Prestressed Concrete Members
Agus Maryoto and Takumi Shimomura
Construction Management
1529 Framework for Quality Improvement of Infrastructure Projects
Abukar Warsame

1540 Reconstruction of an Experts Decision Making Expertise in Concrete Dispatching by Machine
Learning
Mojtaba Maghrebi, Claude Sammut and Travis Waller
Geotechnical and Environmental Engineering
1548 Medium-Deep or Very Deep Disposal of Highly Radioactive Waste?
Roland Pusch, Gunnar Ramqvist, Sven Knutsson and Mohammed Hatem Mohammed
1566 Improvement Properties of Cohesion-Less Soil Using Recycled Bassanite
Masaki Kobayashi, Aly Ahmed and Keizo Ugai
1574 Historical and Future Climatic Change Scenarios for Temperature and Rainfall for Iraq
Saleh Zakaria, Nadhir Al-Ansari and Seven Knutsson
1595 Correlation between the Bearing Capacity of Crude Oil Contaminated Soil of Mgbede and the
Percentage Contamination
Iloeje Amechi Francis

Dec. 2013, Volume 7, No. 12 (Serial No. 73), pp. 1471-1484
Journal of Civil Engineering and Architecture, ISSN 1934-7359, USA

Consideration of Shape Characteristics for Apartments
Incorporating Daylight and Energy Performance
Ji-Eun Lee and Kang Up Lee
Department of Architecture, University of Hanyang, Seoul 133-791, South Korea

Abstract: The objective of this paper is to design units with well-lighted environment and low-energy consumption in the apartment
building. Their daylight and energy performance can be determined by the building shape and orientation. The paper initially
produced the results of illuminations and energy efficiency using the daylight and thermal simulations by Autodesk ECOTECT. It
then provided the comparison on simulation results of two type buildings: Flat-type and L-type apartment. The available options for
the design incorporating the environmental performance have less flexibility in flat-type apartments than in L-type ones. The best
unit in the flat apartment is fixed from -45 to 45 degrees rotation, however, that in the L-type one can change depending on rotating
the building. Moreover, if the upper level units had the appropriate environment, the lower level could have larger window areas in
order to meet those daylight performance needs. Results show that the facade design should have the different window areas
depending on the location of each unit. It can assist in the comfort and low-energy consumption design by using simulation tools that
achieve the more predictable understandings.

Key words: Daylight performance, energy efficiency, passive housing, environmental simulation, building shape.

1. Introduction

1.1 Research Background and Purpose
The structure of a passive house is achieved by
optimizing location, placement and shape of the
building plans. For instance, the use of solar energy is
applied in the best way to heat and light the indoor
space without mechanical or electrical supply. In
contrast, houses with a renewable energy system take
advantage of solar energy to store the energy and
transform it into other forms for producing electricity,
heat for spaces and hot water. The system is required
that actively absorbs natural energy and minimizes the
loss of energy. This effect is associated with the
efficiency of the mechanical systems rather than
architectural design. Therefore, optimizing the
utilization of solar energy is the most important to
improve energy conservation.

Corresponding author: Ji-Eun Lee, Ph.D. candidate,
associate research fellow, research fields: architectural design
incorporating environmental performance, design standards for
the apartment renovation and residential environment
improvement. E-mail: jee@lh.or.kr.
The residential building with pleasant environment
and low energy consumption is designed considering
the correlation between form and energy use. For
instance, the sunlight brought into inside plays a
crucial role in making occupants feel comfortable as
well as decreases the energy consumption. This
concept is related to the governments demonstrating
business for energy conservation, and Seoul is the first
region in South Korea applying the energy
management systems. Thus the amount of energy
consumption will be considered as criteria for judging
the value of the building just like fuel consumption
efficiency of the car. Second, energy consumption
regulations for total amount [1] has been in practice
after February 2013. This document will be presented
as a reference for sale and evaluation of construction
consent targeting apartment complexes with more
than 500 households, a typical size complex. This
research was undertaken for predicting daylight
environment and the energy load of each dwelling and
incorporating this performance into the design
process.
DAVID PUBLISHING
D

1472
The purpose is as follows: First, the daylight
environment is analyzed according to the changes of
floor level and building direction of the major two
apartment shapes: Flat-type and L-type; Second, the
energy consumption is calculated by predicting the
thermal environment by simulation; Finally, the
characteristics of each building shape are analyzed,
and the design method of them was drawn for the
optimal passive houses incorporating the
environmental performance.
1.2 Research Methods and Process
Using simulations from ECOTECT analysis 2011,
the morphological features of various apartments are
drawn by lighting and thermal environment of each
dwelling. The illumination is simulated with the
conditions of both a cloudy and a sunny day and the
distribution is shown on the plan. Energy efficiency is
calculated according to the heating and cooling loads
from the outside temperature by the weather
conditions. This result was then compared according
to the location and direction of each dwelling.
In total, daylight environment varies according to
building direction in two shapes. These results
included the performance of lighting and energy
efficiency in each condition. They also explain the
shape characteristics of passive apartment with a
method to improve the daylight environment of each
dwelling.
2. The Keynote of Passive Design
2.1 Building Shape and Day Lighting
The climate zone of South Korea located at the
latitude 37
o
N belongs to 4 climate zone
(mixed-humid) proposed ASHRAE standard
90.1-2007 [2]. In this humid continental and humid
subtropical climate condition, it is important to use
natural day lighting in the area of the south windows
[3]. The south facing openings absorb the morning
sunlight and more direct solar energy compared to
other directions. A south facing space should be
designed as the concept of open space planning. Thus,
the sunlight enters deeply and widely inside the home,
and air, heat and natural light are well circulated.
Finally, the indoor temperature of the winter would
rise by the natural solar energy, even though summer
cooling load might be overloaded. Therefore, the depth
of the room, the south-facing window area ratio and
other conditions should vary with the building shape.
2.2 Window and Shading Design
Window design should reduce heat loss [4] and
enhance the natural light. However, excessive
overheating inside space should be limited. In addition,
the lighting distribution needs to keep a balance
between the south-facing window and all but south
facing window. The illumination and energy
efficiency of window are different depending on the
U-value, window mullion detail, transmittance and the
SHGC (solar heat gain coefficient) [5]. Therefore, all
the windows on each side are designed to the optimal
condition considering natural environment. For
example, a SHGC of 0.719 in northwest Calgary of
Canada is advisable to obtain high solar energy for the
south-facing window. In addition, a U-value of 0.8
W/m
2
K is recommended [6, 7]. These factors
optimize the daylight and solar heat. On the other
hand, the north, east and west-facing window should
have the thermal insulating value (R-7.7) to decrease
the solar heat and install a low SHGC (0.477) to
reduce the absorption of the sun energy to be
overheated by the low altitude [7]. In addition,
maximizing the free cooling generates significant
difference in energy consumption to minimize for an
air conditioning unit. Therefore, a design strategy
such as operable windows and high ceilings should be
optimized for the natural convection, air circulation
and cross ventilation [8].
2.3 Passive House Design
The passive housing design uses solar energy as
integral design criteria. The passive house in South

1473
Korea can be more efficient use of solar energy than
one first created in Germany, due to the lower latitude.
The window design is important for passive house
design to correspond to the daylight characteristics of
each direction. In addition, the thermal performance of
building envelopes such as walls should be optimized
because a different performance is required depending
on the climate [9].
At this time, heating load is five times higher than
cooling load in South Korea, and plans should be
designed so that as much sun energy as possible is
received, heat can be easily transferred to open spaces
while still preventing overheating. In the case of the
north window which has relatively poor condition, the
required degree of the heat loss and the rooms
sunlight should achieve a balance. Securing the air
tightness for the windows and walls can also reduce
the heat loss. Accordingly, heat fluctuation decreases
and thermal comfort increases. In addition, the
environment to minimize use of air conditioning units
and heating energy usage should be kept.
3. The Method for Performance Evaluation
3.1 Simulation for Evaluation of the Lighting
Environment
Two methods both DF (daylight factor) and LEED
standard for lighting were used to measure the
adequacy of the inside illumination changes
depending on the weather [10]. DA (daylight
autonomy)
1
in the typical two standards of the
dynamic simulation was considered. However, UDI
(useful daylight illuminance)
2
was not selected
because setting the operating hours and the adequate
illumination to work is more important factor for
office building than housing [11-13].
First, as the evaluation standard of day lighting for

1
DA defined here is the ratio of the occupied period of time to
satisfy a minimum illumination with only daylight. DA
standards 500 lux longer hours from 7:00 am to 5:00 pm, and
measuring the percentage of time of one year.
2
UDI is a modification of daylight autonomy. This sets three
illumination ranges as useful values: 0-100 lux, 100-2,000 lux
and over 2,000 lux.
LEED, the lighting environment was evaluated with
both daylight factor in the overcast condition (Green
building rating system version 2, 2005) and the
illuminance of solar noon
3
in the clear day of March
or September 21 now in use (LEED EQ 8.1 rev2,
2009). Daylight factor is a measure of the internal
natural daylight levels of overcast conditions for one
year by ECOTECT. The illumination on September
21 in a clear day is measured by using the radiance
linked from it. The reason to be used in conjunction
with ECOTECT and radiance is that DF has no
change depending on the direction because it has the
standard of the overcast condition, i.e., cloudy day.
On the other hand, the change according to the
azimuth can be drawn on a sunny day by the radiance
simulation. This research set the condition at 21
September solar noon (PM 12:26). Its result might be
limited to describe the characteristics of the whole.
However, it can be expected that the sun position of
the autumnal equinox is close to the average position
since the solar altitude is located in the middle of the
point between the highest and the lowest season. In
addition, it is used for measuring sunshine
environment based on a clear day in LEED
certification. The spacing of the measurement points
was 60 cm which can obtain trustworthy results with
the actual illumination [14, 15].
3.2 Simulation for Evaluation of the Energy
Consumption
Despite energy plus makes it possible to provide the
more realistic results [16], Autodesks ECOTECT
2011 was used to evaluate the energy consumption.
The reason is that it is better for calculating the
amount of electricity rather than heating and cooling
loads. This simulation is faster and easier so it fits the
purpose of this paper to follow a relatively simple
process for reviewing the building forms at the
schematic or design development stage. However, the

3
Solar noon: the point that the altitude of the sun is the highest.
September 21 in Incheon, 12:26.

1474
measurement of the energy efficiency using it creates
a lot of difference depending on setting the material
property such as wall. Therefore, finding the optimal
form is aimed at relative comparisons rather than
absolute ones. The energy efficiency measures the
heating and cooling loads, and the machine energy
consumption of the home is not considered because a
large part of it is determined by the performance of
the machine rather than the architectural design. In
addition, the energy produced by the renewable
energy is not considered even though it is essential
elements for aggressive energy-saving for the same
reason [17].
4. The Environmental Performance
Evaluation of Various Apartment Types
4.1 Modeling Overview
The architectural modeling was carried out by
ECOTECT and set conditions such as weather and
wall materials. Building modeling was performed
based on the shape and area of typical types in the
actual plan drawings and competition drawings. The
two types: Flat-type and L-type were compared as
South Koreas representative apartment types. The
flat-type apartment was common until the late 1990s,
and since 2000, various L-shape or Y-shape tower
apartments have developed. In addition, there are also
mixed type, both flat-type and tower-type. In addition,
the three different floors: 1st floor (bottom), 10th floor
(middle) and the 20th floor (top) were evaluated in the
20 stories building and Table 1 is an overview of the
target model. The weather condition is based on
Incheons data provided in Energyplus weather format
[18].
The material properties set the characteristics of the
buildings exterior walls. In walls, the U-value of 0.30
W/m
2
K matched less than the criteria of the Central
Region (0.47 W/m
2
K) recommended by the Ministry
of Land, Transport and Maritime Affairs in South
Korea. In windows, the Incheons guideline the
guidelines for green building design for a low-carbon,
green growth was referred for deciding the property
[19]. According to these, U-value of window sets 1.7
W/m
2
K, which in Incheon is less than 1.8 W/m
2
K
and is on the basis of more than 2 grade in energy
efficiency rating of windows and doors (1.4-2.1
W/m
2
K). SHGC was 0.81 and visible light
transmittance was set to 0.647. In the use of scenarios,
there were four people usually sitting down and a
mixed-mode system for heating and cooling of inside.
An overview of the applied system of the building is
seen in Table 2.
4.2 Setting the Measurement Condition
The apartment as a single building without
neighboring buildings is not a realistic situation. In
reality, most buildings are influenced by the shadow
of another building. Most of cases have the negligible
impact on the illumination and in some cases a critical
impact on it [20, 21]. However, the placement of

Table 1 Overview of modeling.
Flat-type L-type
Building shape

Floor area All : 129.3 m
2
A, D: 132.9 m
2
B, C: 128.3 m
2

Height/stories 2.8 m/20 stories 2.8 m/20 stories
Total window area 26.4 m
2
A, D: 26.5 m
2
B, C: 25.6 m
2

Window to floor area ratio 20% 20%
South window to wall area ratio
43.7%
(wall: 38.9 m
2
, window: 17 m
2
)
A, D: 39.6% (wall: 22.3 m
2
, window: 8.8 m
2
)
B, C: 34.5% (wall: 63.1 m
2
, window: 21.8 m
2
)


1475
Table 2 Overview of the applied system.
Classification Contents
Wall-skin U-value
Wall 0.30 W/m
2
K
Window 1.7 W/m
2
K (SHGC: 0.81)
Slab on ground 0.35 W/m
2
K
Slab between floors 0.58 W/m
2
K
Roof 0.206 W/m
2
K
Internal design condition
Clothing 0.4 clo
Humidity 60%
AIR speed 0.5 m/s
Lighting level 500 lux (working plane: 70 cm)
Infiltration rate
Air change rate 0.25 ach
Wind sensitivity Reasonably protected (0.25 ach)
Active system Mixed-mode system
Lower band: 18
Upper band: 26
Hours of operation: 24 h

the building would vary depending on the building
shape or the area shape. Hence, the scope of the study
was limited for these reasons, the purpose of the study
is to derive the characteristics of each building
regardless of shading effects due to nearby buildings.
Thus the daylight conditions of each dwelling were
analyzed depending on various building orientation.
Five different angles of rotation axis were chosen, as
shown in Table 3: 0
o
(south-facing), N15E (-15),
N30E (-30), N45E (-45) and N45W (45). The
weather data in Incheon provided by Energyplus was
used for analysis of illumination and energy
efficiency.
4.3 The Simulation of Illumination
4.3.1 The Flat Type Building
The useful illumination has to be more than 2% of
daylight factor [22] and the shading is needed in more
than DF 5% due to glare and solar gain. Table 4
shows that the side dwellings of flat-type building

Table 3 Overview of the measurement point (September 21, 9:00-17:00).
Flat type L type Rotation of axes
Points

Table 4 Daylight simulation of the flat-type apartment.
Location of units Floor No. DF (%) DA (%)
Illumination (September 21 PM 12:26)
N45W 0 N15E N30E N45E
W Unit
(W-west)
1F 4.85 82.28 4,303.85 4,831.48 4,656.14 4,653.93 4,076.84
10F 5.58 86.02 4,437 5,034 4,843 4,823 4,201.39
20F 5.69 86.56 4,688.5 5,410.5 5,196.56 5,141.4 4,462.9
M Unit
(M
-middle)
1F 3.63 61.73 4,040.51 4,791.71 4,705.88 4,713.03 4,143.12
10F 3.84 67.35 4,195.96 5,034.13 4,957.37 4,915.71 4,289.59
20F 3.87 67.52 4,471.1 5,419.26 5,293.43 5,236.62 4,548.28
E Unit
(E-east)
1F 4.02 71.25 4,131.34 4,779.73 4,601.29 4,786.79 4,131.76
10F 4.53 79.02 4,271.32 5,025 4,850.85 4,976.89 4,281.14
20F 4.58 79.29 4,540.47 5,388.61 5,186.17 5,311.43 4,533.76


1476

have better distribution. To be specific, since
dwellings with higher daylight factor are located on
the west more, the highest internal natural lighting
level was likely to be located from the west. However,
illuminations for September 21 at solar noon were
roughly alike regardless of the dwellings different
location.
As shown in Fig. 1, according to floor level, the
difference of DF between 10th and 20th floor dropped
sharply as its 9.8-15% between 1st and 10th one. In
addition, the DAs one did its 3-14% between 1st and
10th

one. The illumination for September 21 with
from -30 to -45 degrees declined sharply when
compared with the rotation angle of -30-0 degrees.
That is the difference between -45 to -30 degrees in
1st floor increased by 3.25, 7.24 and 3.53 times more
than it of -30~0
o
, respectively. On the other hand, the
difference between 45 and -45 degrees is not that big.
The authors can know that the illumination between
east and west with the same angular size is analogous;
that is, the distribution of lighting is expected to be
symmetrical arrangement. As a result, the illumination
of 1st floor would have the poor condition if the
window area is the same as that of upper level. In
addition, it would be better that south-facing buildings
rotate less than 30 degrees. However, if rotating the
building by over 30 degrees were so inevitable,
enough window area should be installed.
4.3.2 The L-Type Building
Each dwelling had relatively small DF gap. To be
specific, the east-locating units (A and B) had the
smallest one. Side units (A and D) have relatively
high DF than middle units (B and C). As shown in
Table 5, in the change of DF from 1st to 20th floor, A,
B and D almost scaled 8% up and C formed a nearly
horizontal line. A or B unit increased rapidly on 10th
floor while average increase of around 8%. In addition,
the changing amount of DA is similar to that of DF.
However, considerable differences of the
illumination at September 21, solar noon on a sunny
day existed between units. For instance, according to
the building rotation, B and C unit totally have higher
illumination and big changing range. Moreover, in A
and B unit, 45 degrees of the rotation angle was the
lowest point and it gradually increased to the point of
minus 45 degrees; in C and D unit, the illumination
data decreased from minus 45 degrees contrary to
previous results as given in Fig. 2.
4.4 Simulation Related to Energy Efficiency
4.4.1 The Flat Type Building
Heating and cooling loads decide consumption of
thermal energy as a big part of residential buildings.
Table 6 shows energy loads of flat type apartment. As

Fig. 1 Distribution of daylight in the flat-type apartment.
6,000
5,000
4,000
3,000
2,000
1,000
0 I
l
l
u
m
i
n
a
t
i
o
n

0
9
1
2

s
o
l
a
r

n
o
o
n

(
L
u
x
)

D
a
y
l
i
g
h
t

a
u
t
o
n
o
m
y

(
%
)

1477
Table 5 Daylight simulation of the L-type apartment.
Location of units Floor No. DF (%) DA (%)
Illumination (September 21 PM 12:26)
N45W 0 N15E N30E N45E
A Unit 1F 4.99 79.62 818.53 1,794.85 2,099.2 3,034.78 2,907.85
10F 5.54 82.75 865.64 1,908.63 2,248.8 3,213.63 3,087.19
20F 5.8 84.7 834.52 1,835.48 2,163.28 3,109.49 3,006.84
B Unit 1F 4.84 78.7 1,892.05 4,225 5,937.26 6,204.23 6,356.21
10F 5.49 83.44 1,997.05 4,500.18 6,308.7 6,618.29 6,770.44
20F 5.61 84.01 1,939.76 4,353.05 6,104.03 6,393.25 6,552.23
C Unit 1F 4.77 76.76 6,745.26 4,760.36 3,927.12 2,165.39 1,985.31
10F 4.77 75.85 7,149.03 4,851.06 3,942.49 2,096.44 1,875.7
20F 4.85 76.24 6,881.21 4,677.39 3,886.37 2,201.78 1,825.61
D Unit 1F 5.56 83.94 3,024.7 1,970.32 1,664.45 932.06 854.08
10F 6.02 85.78 3,086 1,921.15 1,691.66 954.9 873.84
20F 6.53 88 3,113.13 2,011.98 1,691.9 939.91 868.14

Fig. 2 Distribution of daylight in the L-type apartment.

can be seen, the total loads of middle story tends to be
lower and those of the bottom or top floor facing on
the outside are higher. In the heating load on the 1st
floor, there was a significant difference between east
and west unit (W
0
= 63,782; E
0
= 60,732), however,
there was a slight difference in that of 10th or 20th
floor (10F: W
15
= 61,843; E
15
= 62,055). In addition,
the difference between E and M unit in 0 degrees is 4
times more than it between W and M unit, whereas, in
20th floor it is reversed. The -15 degrees among five
cases was the lowest energy load, which is related to
the optimal minus 5 degrees (N5E) in the climate
condition of Incheon with ECOTECT simulation.
When the angle from minus 15 degrees increased or
decreased, total load of units was mostly higher as
shown in Fig. 3. In addition, the heating load of the
middle story is relatively lower because top and
bottom floors face on the outdoor air instead of other
dwellings, however, this factor is not absolute, since
the insulation design method of bottom and top slab is
most important valuable of setting the difference of
energy performance.
4.4.2 The L-type Building
Cooling load remained more or less constant even
though the azimuth angle was changed, however,
heating load had more significant change. Energy
loads of the east dwellings (A and B) were smaller
than those of the west ones (C and D) as shown in
Fig. 4. The case of B unit showed the highest direct
solar ratio and the relatively lower energy loads in total
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
I
l
l
u
m
i
n
a
t
i
o
n

0
9
1
2

s
o
l
a
r

n
o
o
n

(
L
u
x
)

D
a
y
l
i
g
h
t

a
u
t
o
n
o
m
y

(
%
)

1478

Table 6 Energy simulation of the flat-type apartment.
Location Story Azimuth N45W 0 N15E N30E N45E
W
1F
Total (wh/m
2
)
Heating
Cooling
68,252
65,191
3,061
66,616
63,782
2,834
66,521
63,688
2,833
66,741
63,883
2,858
67,228
64,315
2,913
Direct sol (%) 15.7 16.9 17.1 17.2 16.9
10F
Total (wh/m
2
)
Heating
Cooling
62,569
59,188
3,380
60,933
57,763
3,170
60,847
57,676
3,172
61,096
57,880
3,216
61,583
58,312
3,271
Direct sol (%) 16.1 17.3 17.6 17.6 17.3
20F
Total (wh/m
2
)
Heating
Cooling
66,630
63,296
3,334
64,991
61,843
3,148
64,900
61,756
3,144
65,131
61,953
3,178
65,629
62,390
3,240
Direct sol (%) 15.9 17.1 17.4 17.5 17.1
M
1F
Total (wh/m
2
)
Heating
Cooling
67,916
65,441
2,476
67,155
64,832
2,324
67,123
64,780
2,342
67,296
64,880
2,416
67,602
65,102
2,500
Direct sol (%) 14.7 14.1 14.4 14.9 15.2
10F
Total (wh/m
2
)
Heating
Cooling
62,285
59,027
3,257
61,515
58,418
3,098
61,494
58,377
3,116
61,671
58,479
3,192
61,968
58,692
3,276
Direct sol (%) 12.8 12.3 12.6 13 13.3
20F
Total (wh/m
2
)
Heating
Cooling
66,347
63,147
3,201
65,596
62,540
3,056
65,577
62,495
3,082
65,742
62,592
3,150
66,036
62,803
3,233
Direct sol (%) 12.7 12.1 12.4 12.9 13.1
E
1F
Total (wh/m
2
)
Heating
Cooling
65,126
62,045
3,081
63,606
60,732
2,874
63,557
60,698
2,859
63,843
60,952
2,891
64,359
61,431
2,928
Direct sol (%) 16 17.3 17.6 17.6 17.3
10F
Total (wh/m
2
)
Heating
Cooling
62,673
59,284
3,388
61,159
57,982
3,177
61,130
57,948
3,182
61,430
58,206
3,224
61,956
58,679
3,278
Direct sol (%) 16 17.3 17.6 17.6 17.3
20F
Total (wh/m
2
)
Heating
Cooling
66,741
63,392
3,349
65,210
62,055
3,155
65,173
62,018
3,154
65,461
62,269
3,193
65,994
62,748
3,246
Direct sol (%) 15.8 17.1 17.4 17.5 17.1

Fig. 3 Variation of thermal load with different azimuth (flat-type apartment).
70,000
68,000
66,000
64,000
62,000
60,000
58,000
56,000
54,000
52,000
T
o
t
a
l

h
e
a
t
i
n
g

a
n
d

c
o
o
l
i
n
g

l
o
a
d
s

(
w
h
/
m
2
)

D
i
r
e
c
t

s
o
l
a
r

r
a
t
i
o

(
%
)

i
n

p
a
s
s
i
v
e

g
a
i
n
s


1479

Fig. 4 Variation of thermal load with different azimuth (L-type apartment).

like middle units of flat type building. In addition, it
demonstrated that DF or DA, i.e., the internal natural
light level was significantly higher, since the
illumination and the amount of solar energy through
windows is enough. It differed very markedly from C
unit with the lowest direct solar ratio. In addition, B
and C unit showed the similar change pattern in which
total load decreased from 45 degrees. On the other
hand, total load of A or D increased gradually. In 45
degrees facing on southwest, the gap of energy loads
between middle (B and C) and the side units (A and D)
was smaller and the value was more than other
rotation angles as shown in Table 7. According to the
rotation of main axis, the unit with the lowest total
load moved to other location because the main
direction of each dwelling changed. Also, middle
story units tend to have lower loads like as in the flat
type building. However, the higher insulation levels
would make up for the weak points.
4.5 The Environmental Performance Evaluation
4.5.1 The Evaluation of Daylight Environment
The difference in the azimuth and the floor level
was analyzed with the sunshine environment
evaluation. First, the changes of daylight factor
showed the fact that they in the flat-type apartment
form lines with the similar slope on the graph as
shown in Fig. 5. The daylight factor actually
represents internal natural light levels regardless of the
changing climates with a great deal of sunlight. On the
other hand, the L-type one showed that the change of
A and B unit located on the east was analogous to the
west unit of flat-type apartment, whereas, that of C
and D unit forms close to a straight line. This
explained that they had a constant ratio depending on
floor levels. Moreover, side units (A and D) had a
bigger change like them in the flat one.
Also, in the illumination on September 21 of the
flat-type, the units had smaller difference according to
the rotation of the building. However, the L-type
apartment made greatly bigger difference of the
illumination on September 21 between units. In
addition, in the change pattern of the illumination,
those of the flat-type had similarities, whereas, those
of the L-type varied depending on the location of units.
This was because all dwellings in the flat-type
apartment have the same azimuth based on the solar
orientation of the building; whereas, those of L-type
one had the various azimuth.
This explained shape characteristic for apartment as
follows: First, the flat type building could be formed
with vastly different conditions. That is the location of
units would decide the favorable environment
regardless of the change of seasons or the building
arrangement; Second, the L-type apartment could
provide the internal natural daylight not differing
70,000
68,000
66,000
64,000
62,000
60,000
58,000
56,000
54,000
52,000
T
o
t
a
l

h
e
a
t
i
n
g

a
n
d

c
o
o
l
i
n
g

l
o
a
d
s

(
w
h
/
m
2
)

D
i
r
e
c
t

s
o
l
a
r

r
a
t
i
o

(
%
)

i
n

p
a
s
s
i
v
e

g
a
i
n
s


1480
much from other units, especially on 1st floor.
However, according to the season, each unit would
have very different daylight environment. Likewise,
the daylight environment of each dwelling was
influenced by the shape of buildings. Thus, it would
change the design method for building skin. These
form-determining factors beyond the window area
would depend on installing the sunshade, the U-value
of windows, SHGC, and the location and shape of
windows.

Table 7 Energy simulation of the L-type apartment.
Location Story Azimuth N45W 0 N15E N30E N45E
A
1F
Total (wh/m
2
)
Heating
Cooling
63,459
60,499
2,960
64,341
61,463
2,878
64,818
61,969
2,849
65,320
62,501
2,820
65,846
63,054
2,793
Direct sol (%) 19.5 17.6 16.5 15.1 13.7
10F
Total (wh/m
2
)
Heating
Cooling
61,033
57,755
3,278
61,955
58,719
3,236
62,434
59,219
3,215
62,925
59,740
3,186
63,443
60,285
3,158
Direct sol (%) 19.5 17.6 16.5 15.1 13.7
20F
Total (wh/m
2
)
Heating
Cooling
64,928
61,577
3,350
65,821
62,523
3,298
66,297
63,019
3,278
66,800
63,539
3,261
67,307
64,085
3,223
Direct sol (%) 19 17.2 16.1 14.8 13.4
B
1F
Total (wh/m
2
)
Heating
Cooling
64,075
59,677
4,398
62,209
57,814
4,395
61,506
57,174
4,332
60,902
56,656
4,246
60,560
56,400
4,160
Direct sol (%) 27 30.4 31.2 31.7 31.8
10F
Total (wh/m
2
)
Heating
Cooling
61,816
57,162
4,654
60,007
55,333
4,674
59,295
54,700
4,595
58,707
54,192
4,514
58,394
53,940
4,454
Direct sol (%) 27 30.4 31.2 31.6 31.8
20F
Total (wh/m
2
)
Heating
Cooling
65,865
61,104
4,761
64,058
59,263
4,795
63,352
58,616
4,736
62,696
58,084
4,612
62,334
57,802
4,532
Direct sol (%) 26.9 30.3 31 31.5 31.7
C
1F
Total (wh/m
2
)
Heating
Cooling
66,828
63,866
2,963
67,053
63,865
3,188
66,775
63,508
3,268
66,351
62,987
3,363
65,819
62,390
3,428
Direct sol (%) 9.9 10.7 12.1 13.8 15.5
10F
Total (wh/m
2
)
Heating
Cooling
64,359
61,066
3,293
64,529
61,062
3,467
64,260
60,708
3,552
63,823
60,202
3,621
63,313
59,619
3,693
Direct sol (%) 9.9 10.7 12.1 13.8 15.5
20F
Total (wh/m
2
)
Heating
Cooling
68,316
64,944
3,372
68,484
64,950
3,535
68,209
64,595
3,614
67,783
64,088
3,695
67,250
63,504
3,746
Direct sol (%) 9.7 10.5 11.8 13.4 15.2
D
1F
Total (wh/m
2
)
Heating
Cooling
64,119
59,889
4,231
65,522
61,380
4,142
65,938
61,816
4,122
66,273
62,169
4103
66,581
62,479
4,101
Direct sol (%) 30.3 27.5 26.6 25.7 25
10F
Total (wh/m
2
)
Heating
Cooling
61,210
56,752
4,458
62,583
58,187
4,396
62,959
58,583
4,377
63,250
58,891
4,359
63,508
59,162
4,347
Direct sol (%) 29.4 26.6 25.7 24.8 24.2
20F
Total (wh/m
2
)
Heating
Cooling
65,539
60,951
4,588
66,934
62,398
4,536
67,344
62,826
4,517
67,672
63,173
4,499
67,969
63,483
4,486
Direct sol (%) 29.9 27.2 26.3 25.4 24.7


1481

Fig. 5 Natural daylight levels on different floor level.

4.5.2 The Evaluation of Energy Consumption
Heating and cooling loads were measured for
prediction of the energy consumption. The ratio of
direct solar gains showed the direct impact of sunlight
as the element affecting the thermal environment.
Simulation results by ECOTECT showed that the
change of energy loads according to the floor level
was relatively bigger than the azimuth. Units in the
middle story have the lowest energy load in common.
In the case of flat building, total load of west units
with minus 15 degrees, i.e., the azimuth 345
o
was the
lowest. In L-type building, the west middle units, B
with minus 45 degrees, i.e., the azimuth 315
o
had the
lowest ones. The likely reason is that the ratio of
direct solar gains is high and the azimuth is optimized
for catching energy from the Sun. In addition, the
results of units in the flat type were symmetric with
respect to the point of 0 or 15 degrees, whereas, there
is a huge difference between east and west units in
L-type one. Moreover, as shown in Table 7 given
previously, there were substantial changes depending
on the building rotation. The result of flat apartment
would be shown as in Fig. 6. First, the energy
changing pattern between units appeared generally
similar. The difference of total load between the east
and west unit was not that big and they had better
condition in the daylight factor. However, the
difference of total load between 1st and 10th floor was
considerable. In this sense, the units under 10th floor
had a relatively poor condition in the environmental

Fig. 6 Variation of environmental performance with different azimuth (flat-type apartment).
70,000
68,000
66,000
64,000
62,000
60,000
58,000
56,000
T
o
t
a
l

h
e
a
t
i
n
g

a
n
d

c
o
o
l
i
n
g

l
o
a
d
s

(
w
h
/
m
2
)

D
a
y
l
i
g
h
t

f
a
c
t
o
r

(
%
)

1482

Fig. 7 Variation of environmental performance with different azimuth (L-type apartment).

performance. The daylight level also had a relatively
big difference between 1st and 10th floor. It showed
that the amount of solar gain between over and under
10th floor was actually different.
Fig. 7 presents the result of L-type apartment with
different azimuth. There were the significant
difference of energy loads depending on the azimuth
and location of each unit. However, the change of
floor level made smaller difference than it in flat-type
apartment. In addition, it did not interrelate with the
change of daylight level which sharply increased from
1st to 10th floor in the A and B unit and incremented
to 20th floor in the C and D unit. As a result, contrary
to popular belief, B unit facing south-west totally had
the best level, while C unit facing south-east to was
the worst part in both daylight and energy-saving.
To achieve the optimal building design, the
condition which direction each unit faces on and how
much the gap between them appears in the same
building should be considered. In addition, the higher
insulation performance is indispensable for decreasing
thermal load, and then proper window area and
shading installation could enhance the energy
efficiency. Likewise, a design method for the
environmental improvement of dwellings with
relatively poor conditions must be carefully
considered.
5. Discussion and Conclusions
This study has confirmed that the morphological
features of the apartment building affect energy
consumption and comfort for occupants. They can be
improved by a variety of conditions such as inside
illumination, the shape and size of buildings and
windows as results of lots of previous researches have
shown. In addition, the environmental difference of
each unit is analyzed by the simulation depending on
the apartment shape. The design with optimized
environmental performance has to follow a process
where the performance could be predicted and
improved by using the simulation. However, the
current simulation tools are limited in that it is
difficult to check how the design results are close to
the actual extent and which conditions are involved
under simulated results. Thus through the comparative
method in two types of the apartment building, the
daylight environment and energy performance of each
unit has been examined.
Conclusions are drawn as follows: First, the optimal
design of the flat type apartment can be fixed more
efficiently, however, that of the L-type one is flexible
depending on the site condition. In the flat type, the
daylight condition shows more consistent pattern of
change depending on the rotation and elevation of the
building. Moreover, in the energy consumption, the
70,000
68,000
66,000
64,000
62,000
60,000
58,000
56,000
54,000
52,000
T
o
t
a
l

h
e
a
t
i
n
g

a
n
d

c
o
o
l
i
n
g

l
o
a
d
s

(
w
h
/
m
2
)

D
a
y
l
i
g
h
t

f
a
c
t
o
r

(
%
)


1483
elevation is more decisive than the azimuth. Units on
the middle story have lower energy loads than them
on the top or bottom floor. Thus each units of middle
story has the lowest energy load with the different
azimuth. To be specific, the middle story with 0-15
o

azimuth has the lowest one. In the internal natural
daylight level, i.e., DF, the higher floor in the building,
the more differences between edge and middle units
happen, while there are small change in the vertical
line at the same location. Therefore, the flat-type
building can set the 0-15
o
azimuth for the optimal
arrangement. On the other hand, in L-type apartment,
the daylight condition changes with different patterns
of each unit according to the rotation and elevation of
the building. In addition, the energy load has more
change amount than that in the flat type according to
the building rotation, whereas, there was a smaller
change depending on the elevation. Therefore, the
various arrangement methods in L-type depending on
the site condition could be considered for the better
condition of all units.
Second, when the buildings rotation angle
increases from -45 to 45 degrees, the best unit in the
flat apartment remains the same, however, that in the
L-type one alters significantly. In flat-type, the energy
loads did not appear differently in line with the unit
location, whereas, the illumination increased in the
unit located on the side and upper part. Therefore,
units in the middle of the flat apartment should have
relatively large window area than the end ones in
order to get enough daylight. In addition, the top or
bottom slab of the building should install the high
performance insulation for reducing the thermal load.
In L-type, the changing pattern of environmental
performance shown in Fig. 7 varies greatly depending
on the units location. The azimuth made results of the
thermal load greatly different. However, there was
smaller difference depending on the floor level in
terms of indoor thermal environment and comfort of
the units. Therefore the horizontal position and the
solar azimuth angle have a decisive effect on the
units environmental performance in L-type building.
In total, the facade design should depend partly on
the result of simulation in order to design the passive
house with optimized environmental performance. If
passive houses were more common in reality and most
design processes were integrated to check and develop
the energy efficiency, errors and complexity of the
simulation could be improved gradually, and the
optimized design be drawn more easily. In addition,
the most common condition is that other buildings
shadow affects illumination and temperature inside
buildings. From here, future lines of research could
both simulate real-world conditions adequately, as
well as understand the effect of building arrangement
on environmental performance. As a result, the current
design problem that the facade of apartments has the
uniformly repetitive patterns would resolve itself. The
cost of construction could increase in this process.
However, if money on their energy bills could be
saved by these designs, it could be a well-founded
investment.
References
[1] Seoul City Hall, Building Energy Consumption
Regulations for Total Amount, http://citybuild.seoul.go.kr
(accessed Mar. 3, 2012).
[2] ANSI/ASRAE/IESNA Standard 90.1-2007, Energy
Standard for Buildings Except Low-Rise Residential
Buildings, 2007.
[3] Windows and Doors Design Guidelines for Building
Energy-Saving, Ministry of Land, Transport and
Maritime Affairs, July 2012, pp. 8-55.
[4] J. Cheol-Yong, H. Hye-Sim, L. Jin-Sook, The building
energy efficiency rating evaluation of apartment
depending on SC and window area ratio, The Korean
Solar Energy Society 30 (5) (2010) 38-43.
[5] A. ALAnzi, D. Seo, M. Krati, Impact of building shape
on thermal performance of office buildings in Kuwait,
Energy Conversion and Management 50 (3) (2009)
822-828.
[6] Passive Solar House Design, CMHC (Canada Mortgage
and Housing Corporation), Feb. 2012.
[7] W. Feist, Passive House Planning Package 2007-2nd
Revised Edition of English PHPP 2007 User Guide,
Passive House institute, Mar. 2010.
[8] K. Yeang, Designing with Nature: The Ecological Basis

1484
for Architectural Design, Mc Graw-Hill, USA, 1995.
[9] S. Lang, Energy and environmental of residential
buildings in China, Energy and Buildings 36 (12) (2004)
1191-1196.
[10] C.F. Reinhart, J. Weinold, The daylighting dashboardA
simulation-based design analysis for daylit spaces,
Building and Environment 46 (2) (2011) 386-396.
[11] C.F. Reinhart, D.A. Weissman, The daylit
areaCorrelating architectural student assessments with
current and emerging daylight availability metrics,
Building and Environment 50 (2012) 155-164.
[12] A. Nabil, J. Mardaljevic, Useful daylight illuminances: A
replacement for daylight factors, Energy and Buildings
38 (7) (2006) 905-913.
[13] K. Dong-Hwan, Analysis of useful daylight illuminance
(UDI) by dynamic daylight simulation using weather data,
Journal of Architectural Institute of Korea 26 (6) (2010)
321-331.
[14] F. Chella, Compared numerical and reduced scale
experimental analysis on light pipes performances, in: 5th
International Conference on Sustainable Energy
Technologies, USA, 2006.
[15] C.F. Reinhart, O. Walkenhorst, Validation of dynamic
RADIANCE-based daylight simulations for a test office
with external blinds, Building and Buildings 33 (7) (2001)
683-697.
[16] S.R. Chinnayeluka, Performance assessment of
innovative framing systems through building information
modeling based energy simulation, Master Thesis,
Osmania University, 2011, pp. 32-71.
[17] A. Kwang-Ho, K. Hyung-Keun, C. Yong-Seok, A Study
on energy conservation plan of eco-friendly school by
energy plus, Journal of Architectural Institute of Korea 27
(12) (2011) 19-26.
[18] Department of energy, Energy Plus,
http://apps1.eere.energy.gov/buildings/energyplus
(accessed Jan. 1, 2013).
[19] Guidelines for Green Building Design for a Low-Carbon
Green Growth, Incheon City, Apr. 2010.
[20] L. Jang-Bum, Study on the application methods of the
various types of apartment buildings based on the sun
shade layout and the possibility of the advanced
application to PUD planning, Journal of Architectural
Institute of Korea 27 (2011) 165-176.
[21] K. Moon-Il, A study on the application criteria of
daylight prescription, Journal of Architectural Institute of
Korea 25 (2009) 99-106.
[22] M. Moeck, Y.J. Yoon, Green building and potential
electric light energy savigs, Journal of Architectural
Engineering 10 (4) (2004) 143-159.

J ournal of Civil Engineering and Architecture, ISSN 1934-7359, USA

Performance Evaluation of New Two Axes Tracking
PV-Thermal Concentrator
Luis C. Chea
1
, Hkan Hkansson
2
and Bjrn Karlsson
3

1. Department of Physics, Eduardo Mondlane University, Maputo 257, Mozambique
2. Architecture and Built Environment, Lund Technical University, Lund SE-221 00, Sweden
3. School of Sustainable Development of Society and Technology, Mlardalen University, Vsters SE 72123, Sweden

Abstract:

The overall problem with PV (photovoltaic) systems is the high cost for the photovoltaic modules. This makes it
interesting to concentrate irradiation on the PV-module, thereby reducing the PV area necessary for obtaining the same amount of
output power. The tracking capability of two-axes tracking unit driving a new concentrating paraboloid for electric and heat
production have been evaluated. The reflecting optics consisting of flat mirrors provides uniform illumination on the absorber which
is a good indication for optimised electrical production due to series connection of solar cells. The calculated optical efficiency of
the system indicates that about 80% of the incident beam radiation is transferred to the absorber. Simulations of generated electrical
and thermal energy from the evaluated photovoltaic thermal (PV/T) collector show the potential of obtaining high total energy
efficiency.

Key words: Solar concentrating, photovoltaic/thermal systems, tracking concentrating systems.

1. Introduction

The increasing demand of energy, the depletion of
fossil fuel and the environmental problems associated
with the use of fossil fuel make the deployment of
alternative renewable energies attractive. Solar
photovoltaic energy is one of the promising forms of
renewable energy.
The high cost of the solar cells is the main problem
limiting the wide utilisation of PV (photovoltaic)
technology in the developing countries. One possible
solution to the high costs is to concentrate irradiation
on the PV modules in order to minimize the required
cell area for the same output. However, with increased
irradiation concentration, there will be a need for
increased cooling of the PV cells in order to lower the
working temperature, prevent damage and conserve
cell efficiency [1]. Usually, a photovoltaic/thermal
(PV/T) concentrating system has a photovoltaic
module cooled by air or water. Hence, not only

Corresponding author: Luis C. Chea, M.Sc., research field:
solar energy. E-mail: luis.chea@uem.mz.
electricity is generated from the absorber but also heat
can be utilized that would otherwise be wasted [2].
PV/T systems where heat also is utilized to provide a
higher total energy output than PV modules and could
be cost effective if additional cost of the thermal unit
is low [3].
Detailed theoretical models of PV/T collectors
showing how the outputs are limited by different
factors have been developed [4]. The major challenge
in designing a concentrating photovoltaic system is to
achieve an acceptable radiation flux distribution [5].
Parabolic reflectors often cause high local irradiance
and cell temperature resulting in panel power loss [6].
The effect of optical imperfections on the array
performance is also discussed by Anton et al. [7].
Three-dimensional concentrating systems require
sun tracking within a narrow angle cone to work by
directing the direct sunlight to the reflecting optics. In
low concentrating applications, a portion of diffuse
light from the sky can also be captured [8]. The
physics behind the concentrating solar systems
requires that tracking accuracy increases as the system
DAVID PUBLISHING
D
Performance Evaluation of New Two Axes Tracking PV-Thermal Concentrator

1486
concentration ratio increases [9]. In high
concentrations, the tracking accuracy should be in the
0.1
o
range to deliver approximately 90% of the rated
power output, while in low concentrating systems it
should be in the 2.0
o
range [10].
In the present work, we evaluate the tracking
capability of two-axes tracking unit coupled to a
concentrating PV/T system installed at Lund
University. The optical system of the collector was
analyzed and the electrical energy and the heat that the
system would produce is simulated for three different
sites.
2. Description of a Two-Axes Tracking
Hybrid Concentrator
The Swedish Company GSE (Global Sun
Engineering) has developed a rather unique 2-axis
tracking and concentrating hybrid based on flat
mirrors named Matarenki Light. The hybrid is
constructed around 2 9 flat mirrors in a parabolic
geometry (Fig. 1a). The mirrors work constructively
acting as a doubly curved shell together with the steel
frame (Fig. 1b). The mirrors surface has its largest
projected width 3.0 m and height 1.7 m. The distance
between the mirror halves is 0.48 m and the focal
length of the paraboloid is 1.47 m. The absorbers are
16.5
o
inclined towards each mirror array.
Solar cells connected in series demand a uniform
irradiation distribution and this is met using flat
mirrors in this design. The absorber is made from two
parallel sheets of stainless steel to get an even
dissipation of heat. The absorber is isolated on the
back side and covered with a front glass. Silicon cells
are not yet laminated on the absorber.
A Wattsun AZ 125 two axes tracking unit [11]
meant for flat solar panels is used for the tracking
system. Azimuth tracking is made by rotating the
reflectors around the centre pole, while the tilt is given
by the elevation actuator. The tracker is powered by
24 V DC. Light sensors are placed on the four sides of
a box and opposing sensors receive equal amount of
brushing light from the direct component of the light
from the sun when tracking is reached. An additional
sensor on top of the box senses the dusk so the
tracking can return to an east position after dark
(Fig. 2).
2.1 Geometry and Light Distribution
Conventional two dimensional concentrating
systems are constructed around a parabolic trough and
three dimensional systems are built around parabolic
dishes. This system is made up of flat mirrors in facet
mounting. Fig. 3 shows the overall mirror dimensions
at the centre of the mirror.

(a) (b)
Fig. 1 Two axes tracking and concentrating PV/T: (a) front view of the collector with one of the absorbers mounted; (b)
backside of the collector showing how it is connected with the steel frame.

1487

Fig. 2 Box with light sensors.

388
388
415
440
440
440
462 370
415
5
5
2
5
8
5
5
0
5
5
0
6
5
0
6
6
0
0
6
0
0
5
8
5
5
5
2

Fig. 3 Picture taken from the double focal length (without
absorber).

To analyze the light distribution over the absorber,
all mirrors were covered except one that gives the
image on the absorber. Then the image was
photographed and the process repeated for all mirrors
in the right mirror set (Fig. 4).
In Fig. 4a, the uncovered mirror reflects the blue
sky, whereas in Fig. 4b, the uncovered mirror reflects
a white cloud. The mirror in the middle row left is
always shaded by the absorber.
2.2 Estimation of Concentration Factor of the System
Fig. 5 shows the shaded areas on the absorber from
the irradiation for all individual mirrors based on
photos and seen from the sun as in Fig. 3. The central
part of the absorber is irradiated with contribution
from the direct solar radiation multiplied with the
cosine of incident angle from each individual 8 active
mirrors. The values are shown in Fig. 5. This gives a
geometrical concentration factor for the central parts
of the absorber, assuming ideal mirrors and it can be
estimated according to:
( )
=
i
C u cos (1)
where, C is the geometrical concentration factor and u
i

is the angle of incidence in each active mirror. From
Fig. 5, the concentration factor will be:
1 +0.96 +2 0.91 +2 0.95 +2 0.92 =7.52.
The ASTM G173 standard [12] gives the total and
direct spectral irradiation for standardized tilted
surface and is the spectral reference for the definition
of solar cell efficiency. The spectral distribution for
the direct irradiation is shown in Fig. 6. Minor
contribution in the 2,500 nm to 4,000 nm range is not
shown in the diagram.
Standard glass with iron content is used for the
mirrors and as a cover for the absorber. The
transmission trough the 4 mm glass mirrors and the 2
mm glass cover corresponds roughly to the
transmission trough a 10 mm single pane window.
The measured spectral transmission for 10 mm glass
taken from Parasol building simulation program [13]
is multiplied with the spectral direct irradiation and
gives 643 W/m
2
integrated over the whole range of the
solar spectrum. The direct integrated irradiation is 900
W/m
2
giving a reduction factor of 0.71 for the
influence from the glass for solar heat.
The efficiency of a solar cell is defined in the
standard with the direct and the diffuse irradiation of
1,000 W/m
2
. With an assumed sensitivity for Si, a
spectral calibration curve was calculated to give the
electrical output for the desired efficiency. This
spectral calibration curve can then be used to calculate
the electrical output for the actual direct spectrum in
the concentrator with ideal mirrors alternatively with
the influence of glass (Fig. 6).

1488

(a) (b)
Fig. 4 Measuring the contribution of individual mirrors on the absorber: (a) left mirror is active in top row; (b) middle
mirror is active in top row.

Incidence
angle 18
cos 18
=0.95
Incidence
angle 18
cos 18
=0.95
Incidence
angle 23
cos 23
=0.92
Incidence
angle 23
cos 23
=0.92
Incidence
angle 0
cos 0
=1
Shaded
mirror
Incidence
angle 16
cos 16
=0.96
Incidence
angle 25
cos 25
=0.91
Incidence
angle 25
cos 25
=0.91

Fig. 5 Drawing of irradiated areas on the absorber from
individual mirrors based on photos, the incidence angle
from the mirror on the absorber is specified.

Similarly the integrated value of transmission
weighted against an assumed sensitivity of Si gives
162 W/m
2
and the reduction factor is 0.72 for the
influence from the glass for solar electricity.
For this type of design, inevitable triangular shaded
areas are obtained. From Fig. 5, 2% of the absorber
area is estimated to be shaded in this way. Most of the
triangular part of shading in Fig. 5 is assumed to be
reduced by adjustments.
The absorber area is 440 mm 530 mm giving an
area of 0.24 m
2
. An active aperture area for the two
halves is 2 0.24 7.52 0.98 =3.54 m
2
.
2.3 Evaluation of Tracking Capability
An additional angle sensor only sensitive in a
narrow angle cone was developed to analyze the
tracker performance in the vertical and horizontal
directions. It is shown in Fig. 7, for each direction, a
thin aperture slot is placed at the top of the box
generating a light spot on a PSD (position sensitive
device) at the base of the box. It consists of a stretched
out photodiode with a common cathode and separate
anodes at each end. By comparing the two currents
produced by the small spot of light, a position of the
spot and hence the incidence angle can be calculated.
For small angles, the deviation angle, A, in degrees
is given by:
2 1
2 1
180
2 I I
I I
d
L
+
|
.
|
\
|
= A
t
(2)
where, L is length of PSD sensor, d is the distance
from the sensor to the aperture, and I
1
, I
2
are the
currents from each end of the photodiode.
Accordingly, horizontal angular deviation A
h
and
vertical angular deviations A
v
can be calculated.
Additionally, a mechanical angle sensor consisting
of a potentiometer was mounted so that the rig
elevation angle was recorded. This means that the
tracking capability could be controlled.

1489

Fig. 6 The spectral influence from 10 mm glass: (a) direct irradiation; (b) direct irradiation through 10 mm standard glass;
(c) spectrally differentiated output from a solar cell in direct solar irradiation; (d) spectrally differentiated output from a
solar cell in direct solar irradiation through 10 mm glass.

(a) (b)
Fig. 7 View of an optical sensing box: (a) with the PSD sensors at the base of the box; (b) the complete angle deviation
sensor.

3. Experimental Results and Simulation of
Energy Output
3.1 Experimental Results
Fig. 4a shows that the shaded mirror is just due to
the absorber outline indicating a correct size of the
absorber. If the dimensions of the absorber exceed the
design dimensions, three mirrors would be shaded
causing substantial drop of the reflected light.
Measurements of incident angle given by individual
mirrors on the absorber and shown in Fig. 5, enabled
to estimate the geometrical concentration, which is
S
p
e
c
t
r
a
l

i
r
r
a
d
i
a
n
c
e

(
W
/
m
2
)
/
n
m

1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
200 300 400 500 600 700 800 900 1,000 1,100 1,200 1,300 1,400 1,500 1,600 1,700 1,800 1,900 2,000 2,100 2,200 2,300 2,400 2,500
Wavelength (nm)
Direct irradiation
Transmission through glass
Output direct
Output through glass

1490
found to be 7.52. The practical concentration factor
was found by calculating the ratio of measured
intensity of solar radiation on the absorber to direct
solar radiation on the collector plan. The averaged
values of solar radiation measured in these two plans
were 4,810 W/m
2
and 811 W/m
2
, giving an average
practical concentration value of 6.0. This means that
the optical efficiency of the collector is 0.797.
From the measurements with photodiodes, we
found the angular deviations from the direction of
light beam, both in horizontal and vertical plan. Fig. 8
shows horizontal angular deviations during a clear day
with occasional clouds. If the value of optical position
providing vertical deviations is added to the value of
the mechanical angle sensor, the solar height is
determined with good accuracy. This is illustrated in
Fig. 9, where theoretical, measured and measured and
corrected solar height is represented. The relatively
small horizontal deviation angles and the good match
between real and measured solar height, indicate that
the tracking unit works fine during clear days.
On partly cloudy days however, unnecessary
energy losses could be expected and added risks for
the durability of the electrical bypass system.
Because of the concentration, the collector must
have a temperature sensor that sends signals to the
controller to defocus the mirrors at excessive
temperatures. This would protect from stagnation
temperatures.
3.2 Simulations of Energy Output
Simulations of the expected thermal and electrical
output for Stockholm, Madrid and Maputo have been
performed. In these simulations, it was assumed that
the system is equipped with solar cells with a
conversion efficiency of 18%. The system optical
efficiency is assumed to be 0.80. Furthermore, it was
assumed that the solar thermal collector has an optical
efficiency of 0.60 and U-value of 1.6 (W/m
2
K),
calculated by the collector aperture area. This means
that 80% of the direct solar radiation on the mirrors
reaches the solar cells and 18% of incident radiation is
converted into electricity and the rest becomes heat,
about 60%. Heat production is estimated at an average
operating temperature of 50 C.
Electrical and thermal energy per unit area is
calculated according to Eqs. (3) and (4):
d d ob b el b el ob el
H K H K E q u q + =

) ( (3)
T U H K H K E
d d ob b th b th ob th
A + =

) ( ) ( u q u q
(4)
In these equations, q
ob
is the electrical/thermal
optical efficiency for beam radiation, K
b
is the

4
3
2
1
0
1
2
3
4
16:45 17:00 17:15 17:30 17:45
D
e
v
i
a
t
i
o
n
i
n
h
o
r
i
z
o
n
t
a
l
d
i
r
e
c
t
i
o
n
]
Time[hh:mm]
Horizontaldeviation

Fig. 8 Angular horizontal deviation of solar tracking during a cloud (hunting event) measured by an optical sensor.
(
o
)

Time (hh:mm)

1491
18
19
20
21
22
23
24
25
26
27
28
16:45 17:00 17:15 17:30 17:45
s
o
l
a
r
a
l
t
i
t
u
d
e
[

]
Time[hh:mm]
theoreticalsolaraltitude
measuredsol aralti tudewithcompensationbyangulardeviationdevice
solartracking

Fig. 9 Vertical tracking for the same event as in Fig. 8: real solar altitude; measured solar altitude and compensate by
vertical deviation; measured solar altitude.

electrical/thermal incidence angle modifier for beam
radiation, defined by Eq. (5), H
b
is the beam
irradiation, K
d
is the incidence angle modifier for
diffuse radiation, H
d
is the diffuse irradiation, U is the
overall heat loss coefficient which is defined by Eq. (6)
and AT is the temperature difference between mean
collector temperature and the ambient temperature.
The incidence angle modifier for beam radiation for
incidence angles between 0
o
and 60
o
is given by:
( ) 1 cos / 1 1 ) (
0
= u u b K
b
(5)
where, b
0
is the electrical/thermal angular coefficient.
The diffuse incidence angle modifier was calculated
as being the inverse of the concentration ratio, and the
U-value is given by:
T a a U A + =
2 1
(6)
where, a
1
is the constant part of the overall heat loss
coefficient, and a
2
is the temperature dependent part
of the overall heat loss coefficient.
Simulations were performed for a flat PV module,
vacuum tube, both at optimum tilt (40
o
for Stockholm,
30
o
for Madrid and 26
o
for Maputo), tracking
concentrating PV module and tracking concentrating
solar collector representing the evaluated PV/T
tracking concentrator. For vacuum tubes, the
parameters are: optical efficiency of 0.80 and U-value
2 (W/m
2
K). Here, we assume that it is a vacuum tube
with flat absorber inside so that single sided absorber
area is used as the defined area.
The results show that the global solar irradiation on
a fixed surface towards equator is higher than the
direct solar radiation on a surface with two axes
tracking and it is independent of latitude. This is
derived from Fig. 10.
A global sun engineering system with an area of
3.54 m
2
provides:
(1) 1,070 kWh/yr of electricity and 3,600 kWh/yr of
heat in Madrid. The system can be replaced in Madrid
by approximately 3.5 m
2
of flat PV modules and 3.5
m
2
of vacuum tubes in order to produce the same
amount of energy;
(2) 600 kWh/yr of electricity and 2,000 kWh/yr of
heat in Stockholm. The system can be replaced by 3
Time (hh:mm)

1492

Fig. 10 Global solar irradiation on inclined surfaces
towards equator, direct irradiation on surfaces with
east-west and two axes tracking in Stockholm, Madrid and
Maputo.

Fig. 11 Electric power delivered by PV modules: inclined
26
o
North, tracking around an east-west axis,
tracking in two axes and concentrating PV plus tracking in
Maputo.

Table 1 Energy output of flat PV, vacuum tubes, concentrating PV and concentrating solar collector in three different
locations.
Stockholm
Vacuum tubes
(kWh/m
2
, yr)
Output concentrated PV
module (kWh/m
2
, yr)
Flat PV module
(kWh/m
2
, yr)
Output concentrated solar
collector (kWh/m
2
, yr)
40 tilt 639 - 195 -
E-W tracking 704 111 212 335
Full tracking 947 152 271 511
Madrid
30 tilt 1,194 - 312 -
E-W tracking 1,296 199 337 606
Full tracking 1,722 268 412 918
Maputo
26 tilt 1,390 - 344 -
E-W tracking 1,494 208 366 698
Full tracking 1,845 266 454 970

m
2
of flat PV modules and 3 m
2
of vacuum tubes;
(3) 1,060 kWh/yr of electricity and 3,880 kWh/yr of
heat in Maputo. The system can be replaced by 3.2 m
2

of flat PV modules and approximately 3 m
2
of
vacuum tubes.
The electrical and thermal output for the three
different sites is summarized in Table 1. For Maputo,
the electrical outputs are represented in Fig. 11.
4. Conclusions
An advanced two-axes tracking solar concentrating
system producing both electricity and heat was
developed by the Swedish company Global Sun
Engineering. Flat mirrors provide uniform light
intensity on the absorber with solar cells. A prototype
of this system has been installed on the roof at Lund
University, aiming to evaluate the complete system.
The evaluation shows that the flat mirrors provide
uniform illumination on almost the entire area of the
absorber. The tracking unit works with good accuracy
on clear days. However, for a partly clouded day,
unacceptable deviation (hunting) occurs. This is
planned to be solved by sensing the currents from
opposing solar cells on the borders of the absorber for
fine tuning of the tracking.
Simulations of the system have been made with
certain assumptions about the system performance.
The results show that:
The global solar radiation on fixed surface
inclined to an optimal angle towards equator is higher
than the direct radiation on a surface with two axes
tracking, in all latitudes;
2,500
2,000
1,500
1,000
500
0
I
r
r
a
d
i
a
t
i
o
n

(
k
W
h
/
m
2
)

E
l
e
c
t
r
i
c

o
u
t
p
u
t

(
k
W
h
/
m
2
)


1493
The analyzed PV/T system with an area of 3.54
m
2
would produce approximately 1,060 kWh/yr of
electricity and 3,800 kWh/yr of heat in Maputo. In
Stockholm, the system would produce approximately
half of these figures. To produce the same amount of
energy, the system can be replaced by 3.2 m
2
of PV
modules and 3 m
2
of vacuum tubes in Maputo.
References
[1] J . Nilsson, H. Hkansson, B. Karlsson, Electrical and
thermal characterization of a PV-CPC hybrid, Solar
Energy 81 (7) (2007) 917-928.
[2] L.R. Bernardo, B. Perers, H. Hkansson, B. Karlsson,
Performance evaluation of low concentrating
photovoltaic thermal systems: A case study from Sweden,
Solar Energy 85 (7) (2011) 1499-1510.
[3] Y. Tripanagnostopoulos, M. Souliotis, R. Battisti, A.
Corrado, Application aspects of hybrid PV/T solar
systems, http://www.enc.nl/fileadmin/ecn/units/egon/pvt/
pdf/eurosun04_lca.pd (accessed Apr. 21, 2010).
[4] J .S. Coventry, K. Lovegrove, Development of an
approach to compare the value of electrical and thermal
output from domestic PV/thermal, Solar Energy 75 (2003)
63-72.
[5] J .S. Coventry, Performance of a concentrating
photovoltaic/thermal solar collector, Solar Energy 78
(2004) 211-222.
[6] I. Anton, R. Solar, G. Sala, D. Panch, Testing of
concentrating modules and cells with non-uniform light
patterns, http://www.ies-def.upm.es/ies/CRATING
/Vc237.pdf (accessed J une 18, 2012).
[7] I. Anton, G. Sala, J .C. Arboiro, J . Mondero, P. Valera,
Effect of the optical performance on the output
power of the Euclides array, in: 16th European
Photovoltaic Solar Energy Conference, Glasgow, 2000,
pp. 2225-2228.
[8] J .A. Duffie, W.A. Beckman, Solar Engineering of
Thernal Process, J ohn Wiley & Son, USA, 2006.
[9] R. Winston, J. Minno, P. Bentez, Nonimaging Optics,
Academic Press, USA, 2005.
[10] Concentrated Photovoltaic Trackers,
http://en.wikipedia.org/wiki/Solar_tracker#Concentrated_
photovoltaic_.28CPV.29_trackers (accessed May 5,
2012).
[11] Wattsun Solar Tracker Installation Guide,
www.wattsun.com (accessed Mar. 10, 2010).
[12] ASTM G173 Standard,
http://pveducation.org/pvcdrom/appendicies/standard-sol
ar-spectra (accessed May 12, 2012).
[13] Parasol Building Energy Simulation Program,
http://www.ebd.lth.se/english/software/parasol/ (accessed
May 15, 2012).

A Study on the Architectural Application of Aerogel
Young Cheol Kwon
School of Architecture, Halla University, Gangwon 220-712, Korea

Abstract: As the building energy saving is more and more important, high-performing insulation like aerogel will be required in
buildings to improve their thermal environment and to save building energy. This study conducted the literature survey on the
aerogel blanket and presents the architectural application considering its high-insulating property. If aerogel is applied to a building
as an envelope insulation, its thickness could be reduced to as half as that of existing insulation such as polystyrene foam board or
glassfiber. Currently, aerogel is largely used as a thermal breaker in thermal bridges in buildings. It is still too expensive to be used
as a main insulation of whole building envelopes. Thanks to the advantages of aerogel blanket such as low thermal conductivity,
broad temperature range for use, excellent water repellent property and fire resistance, easiness of moving and applying, it has much
possibility in the respect of its building application.

Key words: Insulation, aerogel, thermal conductivity, aerogel blanket.

1. Introduction

The depletion of energy resources and the problem
of global warming require people sustainable
development. Sustainability covers many issues, but
none is as important as energy. More than any other
factor, the energy consumption of buildings is
destroying the planet. Buildings use about 40% of all
the energy consumed. This energy is mostly derived
from fossil sources that produce the carbon dioxide
that is the main cause of global warming. People must
replace these polluting sources with clean, renewable
energy sources such as wind, solar energy and
biomass, or people must increase the efficiency of
their building stock so that it uses less energy, or
people must do both. Of course, people need to do
both, but decreasing the energy consumption of
buildings is both quicker and less expensive [1].
To fight global warming, people must understand
what their energy options are. By far the most
important option is efficiency, which is the easiest,
quickest and least expensive way to fight global
warming. The most efficient way to decrease the

Corresponding author: Young Cheol Kwon, Dr., professor,
research fields: building energy saving and insulation. E-mail:
yckwon@halla.ac.kr.
energy consumption of buildings is to increase the
insulation of thermal envelopes such as the roof and
exterior walls.
Aerogel has largely been used as insulation for the
spacecraft and industrial facilities. As the building
energy saving is more and more important,
high-performing insulation like aerogel will be
required in buildings to improve their thermal
environment and to save building energy.
This paper aims to show the possibility of aerogel
as a building insulation.
2. The General Properties of Aerogel
2.1 Aerogel Materials Derived from Polymers Other
than Silica
Essentially, any liquid can be gelled when a
continuous solid lattice structure is formed within the
liquid phase, entraining the solvent within open
pores.
The key to making a nanoporous aerogel involves
developing a balance between polymer chain growth
and interchain cross-linking that will net the desired
pore structure and high surface area. The modulus
(stiffness) of the resulting gels must be high enough to
resist the capillary forces generated during solvent
DAVID PUBLISHING
D

1495
removal. This property is strongly influenced by the
number of links between neighboring chains per unit
volume of gel. Many materials can be incorporated
into the silica- or other metal oxide-matrix at the sol
stage, including organic polymers that uniquely
modify the aerogel physical properties. Many highly
cross-linked polymer systems can also be induced to
create gels in organic solvents, such as
resorcinol-formaldehyde, melamine-formaldehyde,
polyimides, polyurethanes, polyisocyanurates and
various unsaturated hydrocarbon materials.
2.2 Aerogels Have Extreme Structures and Extreme
Physical Properties
The highly porous nature of an aerogel structure
provides a huge amount of surface area per unit
weight. For instance, a silica aerogel material with a
density of about 100 kg/m
3
(or about 1/10th that of
water) can have surface areas of around 800 m
2
/g to
1,500 m
2
/g depending on the precursors and process
utilized to produce it. That is equivalent to roughly 3
to 5 football fields per gram of solid materialshowing
that an extraordinary amount of surface is folded in on
itself within the aerogel structure.
The percentage of open space within an aerogel
structure is about 94% for a gel with a density of 100
kg/m
3
.
2.3 Development of Aerogel Blanket
Aerogel blanket was developed in 1999 by an
American company of Aspen Aerogel. The production
procedure of aerogel blanket is shown in Fig. 1.

Fig. 1 Production procedure of aerogel blanket.

1496

Aerogel production has remained a costly affair
mainly due to high prices of raw materials and
relatively lower production volumes. Traditionally,
aerogel synthesis involves the following three steps
[2, 3]:
2.3.1 Sol-Gel Preparation
The gel is prepared using silicon precursors, such as
tetramethyl orthosilicate (TMOS), tetraethylortho
silicate (TEOS), polyethoxydisiloxane (PEDS),
methyltriethoxysilane (MTES), silicon alkaoxide, etc..
2.3.2 Gel Aging
The gel prepared in step 1 is aged in a solvent for
long periods of time to improve the mechanical and
permeability properties of the gel network. The
concentration, aging time, temperature, pH and
polarity have a strong influence on the strength and
porosity of the aerogel.
In laboratory settings, a batch process is adopted for
the aging step, which is inherently a slow process.
When scaling to production phase, the batch
processing step may increase product cost due to
increased process line downtime and frequent
stopping/starting of the production line. A continuous
process is desirable because it helps to increase output.
However, as with any multi-step process, the
individual step only adds time to the overall process if
it is the bottleneck. Since aging is not the capital
intensive part of the process, larger aging vessels can
be used to reduce the process time for this step.
2.3.3 Gel Drying
In this step, the liquid inside the gel network is
removed using a liquid to gas phase change process.
However, due to the surface tension of the liquid in
contact with the solid, the liquid changing phase tends
to pull the gel network along with it. This causes the
gel network to shrink and collapse. To retain the
integrity of the gel structure during the drying process,
the aged gel is brought to supercritical conditions.
Under the supercritical conditions, surface tension
vanishes as there is no distinct liquid-vapor phase
boundary. The supercritical conditions may be
realized either at low or high temperatures depending
on the liquid (CO
2
, ethanol, etc.), but high pressures
are always required.
As-produced aerogels are fragile and unsuitable for
use in any practical application unless they are
reinforced with glass fiber, mineral fiber, carbon fiber
etc. or cross-linked with polymers [4]. Although the
reinforcement process provides mechanical strength
and flexibility to the aerogel, it may result in an
undesirable increase in the thermal conductivity and
density of the resulting aerogel composite [5, 6]. If the
mechanical requirements for the aerogel application
permit, the thermal conductivity increase might be
minimized by using lower volume fractions of inferior
thermal conductivity fibers. Infrared opacifiers, such
as carbon black, titanium oxide and iron oxide, with
suitable fiber diameters may also be added during the
sol-gel process to reduce the radiative part of the
thermal conductivity [7]. The radiative contribution
can be further reduced by using IR opacified fibers,
such as PET fibers coated with carbon black.
Aerogel blanket is used world-widely especially for
the insulation of spaceships and industrial facilities
thanks to its low thermal conductivity and wide range
of applicable temperature.
Fig. 2 shows the shape of blanket-type aerogel.
2.4 Physical Properties of Aerogel Blanket
The thermal conductivity of aerogel blanket is
0.011-0.015 W/mK, which is half of that of
polyurethane and air [8].

Fig. 2 Aerogel blanket.

1497
Fig. 3 shows the thermal conductivity of aerogel
blanket according to the mean temperature.
Fig. 4 shows the thermal conductivities of various
insulations.
The range of applicable temperature of aerogel
blanket is from -200
o
C to 650
o
C, so it could be used
as a special insulation to meet the needs of both
keeping cold and keeping hot from extreme cold
environment to extreme hot temperature. Fig. 5 shows
the applicable temperature ranges of various
insulations.
Fig. 6 shows it also has the high fire resistance
which can be resistant for the flame of 1,200
o
C.
Fig. 7 shows it has high hydrophobility that prevents
water to be passed but allows air penetration. Aerogel
has the flexibility which allows an easy installation
and has high strength which could be used under the
floor heating pipes like in Fig. 8.
The density of aerogel blanket is 0.10-0.17 g/cm
3
.
The general properties of aerogel blanket can be
summarized as Table 1 [9].

Fig. 3 Thermal conductivity of aerogel according to its
temperature [9].

Fig. 4 Thermal conductivity of insulations

Fig. 5 Applicable temperature ranges of insulations.

Fig. 6 Fire resistance of aerogel blanket.

Fig. 7 Hydrophobicity of aerogel blanket.
Size of a molecule of a vapor: 0.25 nm
Size of an aerogel airspace: 10-50 nm
Size of a water particle: above 100 nm

1498

Fig. 8 Aerogel blanket used under heating pipe.

3. Architectural Applications of Aerogel
Blanket
Up to now, aerogel blanket has usually been used
for the insulation of spaceships or industrial facilities.
However, as the building energy saving is highly
required, it could be effectively used as a building
insulation.
Figs. 9-11 show some architectural applications
[10]. Fig. 9 shows it can be used as an effective
thermal breaker on the framing. Fig. 10 shows its
roofing application. Aerogel blanket is used as an
insulation of envelopes of mobile homes like in Fig. 11.
Aerogel was once introduced as a material to
protect the earth in the future of 2050 through its high
effective insulating performance for building energy
saving through a documentary program of Eco
Polis.
It was reported that If aerogel is used as building
insulation, the green house gas could be reduced as
much as 29%.
While passive house has largely the thickness of
insulation of 25-40 cm for its high insulating
performance, aerogel can meet the same insulating
property with thickness of 9-15 cm [11]. Fig. 12
shows the overall thermal transmittance of thermal
envelopes of a passive house and required thickness of
insulation.
In spite of the incomparable physical performance
of aerogel blanket, the only handicap in its
architectural applications could be its high price.

Table 1 General properties of aerogel blanket.
Thermal conductivity
0.011-0.013 W/m-K at 38 C (100 F) and 760 torr. Conductivity decreases to 0.004 W/m-K at 10
torr.
Constant use temperatures -273 C/-459 F to 650 C/1,200 F
Density Currently available in densities from 0.10 to 0.12 g/cm
3
(6 pcf to 8 pcf)
Pore size Open celled structure (2-50 nm pores) with an average pore size approximately 10 nm
Flexibility Conformable at 0.25 inch thickness, drapeable at 0.125 inch thickness
Compressive strength
Aerogels typically have excellent compressive strength compared to microcellular foams and fibrous
insulations
Hydrophobicity Unencapsulated materials will float on pure water indefinitely and resist liquid water infiltration
Acoustic properties
At 100 m/s, aerogels have an extremely slow speed of internal sound propagation, sound
transmission through aerogel is significantly retarded with fiber reinforcement
Toxicity
Aerogels are based on amorphous silica gel, which is considered safe and non-toxic, in typical
handling, unencapsulated materials will generate nuisance dust

Fig. 9 Aerogel blanket used for breaking thermal bridge through framing.

1499

Fig. 10 Aerogel blanket used in roofing applications.

Fig. 11 Aerogel blanket used in mobile homes.

Overall thermal transmittance of thermal envelopes of a passive
house

Wall 0.09-0.15 W/m
2
K
Floor 0.08-0.15 W/m
2
K
Roof 0.07-0.15 W/m
2
K
Required thickness of insulation for the wall, floor and roof

Styrofoam board 25-40 cm
Aerogel blanket 9-15 cm

Fig. 12 Overall thermal transmittance of thermal envelopes of a passive house and required thickness of insulation.

4. Conclusions
Thanks to the advantages of aerogel blanket such as
low thermal conductivity, broad temperature range for
use, excellent water repellent property and fire
resistance, easiness of moving and applying, it has
much possibility in the respect of its building
application. However, because of its high price, it has
reluctance in the building use.
Therefore, the strategy to establish its priority in
price competition is necessary. Price reduction of the
product and development of complex insulation
system using aerogel blanket are recommended.
Complex insulation system with reflective
insulation could be one of the solutions to secure its
price competition for building application.
It is also recommended to develop a new
not-too-expensive aerogel insulation which has
normal applicable temperature range about from
-50-50
o
C.
Acknowledgments
This work was supported by the National Research
Foundation of Korea (NRF) grant funded by the
Korea government (MEST) (No. 2011-0001361).
This paper is also the result of the research which
was supported by Korea Sanhak Foundation.
References
[1] N. Lechner, Heating, Cooling, Lighting Sustainable
Design Methods for Architects, John Wiley & Sons, New
J ersey, 2009.
[2] N.C. Shukla, A. Fallahi, J . Kosny, Aerogel for thermal
insulation of interior wall retrofits in cold climates, in:
Building Enclosure Science & Technology (BEST3)
Conference, Atlanta, 2012.
[3] A.S. Dorcheh, M.H. Abbas, Silica aerogel, synthesis,
properties and characterization, J ournal of Materials
Processing Technology 199 (2008) 10-26.
[4] Strong and Flexible Aerogels Web Site,

1500
http://www.aerogel.org/?p=1058 (accessed Oct. 8, 2012).
[5] Z. Deng, J . Wang, A. Wu, J . Shen, B. Zhou, High
strength SiO
2
aerogel insulation, J ournal of
Non-Crystalline Solids 225 (1998) 101-104.
[6] Aerogel Web Site http://www.aerogel.com/products/pdf/
Pyrogel_XT_DS.pdf (accessed Oct. 10, 2012).
[7] D. Frank, A. Zimmermann, Aerogel composites, process
for producing the same and their use, US Patent, 5789075
A (1998).
[8] Thermal Conductivity of some common Materials and
Gases Web Site, http://www.engineeringtoolbox.com
/thermal-conductivity-d_429.html (accessed Oct. 10,
2012).
[9] Aspen Aerogels Web Site, http://www.aerogel.com/
(accessed Oct. 10, 2012).
[10] Aerogel Korea Web Site, http://www.aerogel.co.kr/
(accessed Oct. 10, 2012).
[11] Y.C. Kwon, A Study on the Architectural Applications
of Aerogel, Final report, Korea Sanhak Foundation,
2010.


Development of Design Response Spectra Based on
Various Attenuation Relationships at Specific Location
Hapsa Husen
1
, Taksiah Abdul Majid
2
, Fadzli Mohamed Nazri
2
, Mohd Rashwan Arshad
2
, Ade Faisal
2
and Siti
Khadijah Che Osmi
1

1. Civil Engineering Department, Faculty of Engineering, Universiti Pertahanan Nasional Malaysia, Kuala Lumpur 57000,
Malaysia
2. School of Civil Engineering, Universiti Sains Malaysia, Pulau Pinang 14300 , Malaysia

Abstract: Peninsular Malaysia is located and lies in a low seismic region. Although Malaysia is not located in the active fault
seismic area, it is closed to the Sumatran active seismic zones. Tall building are frequently felt the tremor generated from Sumatran
subduction and fault zones especially in the west cost of Peninsular Malaysia such as Johor Bahru, Kuala Lumpur and Penang.
Existing design response spectra was developed based on attenuation relationship for each subduction and fault zone. In this study,
the design response spectra were developed based on various attenuation relationships for selected location in Kuala Lumpur area,
namely, Mutiara Damansara, Bandar Petaling Jaya and Bandar Puteri Puchong. The development of design response spectra based
on various attenuation relationships is more reliable in selecting the appropriate attenuation relationship for the study area. Seven
attenuations have been chosen and results show that Megawati et al. are the most appropriate attenuation relation for fault zone,
where the predicted PGA (peak ground acceleration) is 0.0187 g which is the proposed PGA value for this study area. This study
also found that most of soil in the study area can be categorized into s
D
(stiff soil) according to site classification in the NEHRP
2000 Provision/UBC 97. Bandar petaling Jaya was found to be highest AF (amplification factor) of 3.74 for stiff soil and Mutiara
Damansara with AF of 2.67 for very dense soil or soft rock. The proposed design response spectra for each location were
developed based on UBC 1997 (Uniform Building Code 1997). The peak RSA (response spectrum acceleration) of 0.30 g for soil
type S
D
for Bandar Petaling Jaya is the maximum level of acceleration on the soil surface with a period range of 0.10 to 0.52
seconds. All these values can be used for the seismic safety evaluation of existing structures and as a guideline in designing new
structures to resist future earthquake, within the study area.

Key words: Attenuation relationship, PGA (peak ground acceleration), response spectrum acceleration, response spectra, Sumatra
subduction zone, Sumatra fault zone.

1. Introduction

The Peninsular Malaysia is located in a low-seismic
region where the closest active seismic zones are
located approximately 350 km away that generated
from the Sumatran seismic zones namely Sumatran
subduction and fault zones. Although Peninsular
Malaysia lies on seismically stable region but
structures in this region frequently felt the earthquake
generated from the Sumatran seismic zones as low as
magnitudes of 5.0 [1].

Corresponding author: Hapsa Husen, M.Sc. student,
research field: structural engineering. E-mail:
che_jedi@yahoo.com.
According to Megawati et al. [2], it was predicted
that an earthquake event at a moment magnitude
greater than 7.8 from the Sumatra subduction zone has
capability to generate destructive ground motion in
Singapore and Kuala Lumpur, even at a distance of
700 km.
Pan [3] stated that even a longer distance may create
destructive ground motion, so it seems reasonable to
assume that larger and closer earthquake might results
in tremendous ground motion to Peninsular Malaysia
generally and Kuala Lumpur especially. In addition,
Kuala Lumpur was reported as the most frequent
earthquake felt area in Malaysia by Malaysian
DAVID PUBLISHING
D
Development of Design Response Spectra Based on Various
Attenuation Relationships at Specific Location

1502
Meteorological Services [4].
Kuala Lumpur had rapid development in the past
few decades which is also contribute to increasing of
the construction of tall building or structures, therefore
specific location in Kuala Lumpur area has been
chosen as a case study area in order to carry out the
earthquake study or seismic hazard analysis which is
can be used for future planning especially in safety
design of structure.
This study was carried out to achieve three main
objectives which are to determine and propose PGA
based on various attenuation relationships for study
area, to develop spectral acceleration on the specific
location in Kuala Lumpur, using NERA program and
also to develop design response spectrum using
proposed PGA value and according to UBC 1997, for
study area.
This research is based on the distant earthquake
from Sumatra and limits its scopes of works which are
the maximum magnitude value for subduction zone
and fault zone are taken from Fadzli [5], three
attenuation relationships in determining PGA are
Adnan et al. [6], Lam et al. [7], Megawati et al. [2] and
Petersen et al. [8], for Sumatra subduction and
Campbell [9], Lam et al. [7] and Megawati et al. [10],
for Sumatra fault zones, due to availability soil data
of Kuala Lumpur, only three locations area will be
studied using 28 borelog data, the development of
response spectra acceleration based on local site
condition, one strong-motion data used in establishing
RSA (response spectra acceleration) is El Centro
Earthquake, NERA (non-linear Earthquake site
Response Analysis) software will be used for
performing RSA and design response spectra will be
established based on UBC (1997).
2. Theoretical Background
2.1 General Tectonic Setting of Peninsular Malaysia
Malaysia is situated close to two most seismically
active plate boundaries namely, the inter-plate
boundary between the Indo-Australian Plates and
Eurasian Plates on the west, and the inter-plate
boundary between the Eurasian and Philippines Sea on
the east.
The India-Australia plate subducted below the
Eurasia plate along the arc at a rate of about 67
mm/year and this displacement between the two plates
is partly accommodated by sudden movements which
can cause earthquakes [11].
Generally tectonic features affected Peninsular
Malaysia can be divided by two different
classifications, which is subduction and fault zone. The
nearest distance of earthquake epicentre from Malaysia
is approximately 350 km [6]. Numerous earthquakes
ranging from magnitude 6.5 to 9 have occurred in both
the Sumatran fault system and the subduction
interfaces as reported by Sieh and Natawidjaja [12]
(Fig. 1).
2.2 Sumatran Subduction and Fault Zone
The Sumatra subduction zone is defined as one of
the most active plate in the world tectonic margins
accommodating about 49 to 60 mm/year [10] and has
been the focus of several and large earthquake [8]. The
latest earthquake disaster occurred on the 26 December
2004 that is great Sumatra-Andaman earthquake and

Fig. 1 Active tectonics and seismologic summary of the
Sumatra plate boundary [13].

1503
was the largest seismic event ever recorded with
magnitude of 9.3 that generating tsunami.
Strike slip faults is caused by relative horizontal
displacement of the two sides of faults taking place
along an essentially vertical fault plane, as occurred at
Sumatra great fault zone. The Sumatran faults
ranking as one of the great strike-slip fault on earth,
it is high in level seismicity and it is major role in
active tectonic and seismic hazard of Southeast Asia
[12]. Several large earthquakes have occurred in this
zone as mentioned by Adnan et al. [6] with magnitude
between 6.7 and 7.5.
In this study, the earthquakes sources data from
each Sumatran subduction and fault zone were first
selected where earthquake event with magnitude less
than 5 will be neglected, as suggested by Petersen et. al.
[8] and maximum magnitude values for subduction
and fault zone was taken from study done by Fadzli [5]
where the maximum magnitude for subduction and
fault zones are 9.24 and 7.7, respectively.
2.3 PGA (Peak Ground Acceleration)
PGA is a measure of earthquake intensity.
According to Sokolov [14], the PGA is strictly
depending on earthquake magnitude, distance and
local geological condition. The high peak acceleration
gives more destructive effect than lower peak
acceleration, with percentage of structure damages
depending on the period of PGA. According to Fadzli
[5], the attenuation relations relate the peak ground
acceleration is important parameters in seismic hazard
analysis.
In order to study the effect of earthquake to building
in Malaysia, the first step is to establish the PGA where
the derivations of PGA are based on attenuation
relationship. Adnan et al. [6] stated that the appropriate
attenuation function is the basic fundamental to
perform seismic hazard analysis. Numbers of previous
researcher are developing the response spectral
attenuation relationships for Peninsular Malaysia,
therefore this study was conducted in order to make
suggestion on the best attenuation relationship that
suitable to use in order to develop the design response
spectra for specific location in Kuala Lumpur.
Due to lack of attenuation relations developed for
the distant earthquakes for Peninsular Malaysia, far
distant attenuation functions were chosen in prediction
of PGA value in this study (at study area). According to
Adnan et al. [6], Peninsular Malaysia is affected
seismically by far field earthquake events from
Sumatra fault or Sumatra subduction, therefore the
selection of the attenuation relationship should
consider the effects of far field earthquakes from
Sumatra.
In this study four selected attenuation relations were
used for Sumatran subduction zone namely Megawati
et al. [2], Lam et al. [7], Adnan et al. [6] and Petersen
et al. [8], while Megawati et al. [10], Lam et al. [7]
and Campbell [9] were applied in Sumatran fault
zone.
As suggested by Fahmy [1], comparison between
prediction values of PGA and the measured peak
ground acceleration values recorded by seismology
stations in Malaysia were carried out in this study. This
is because to select the most suitable attenuation model
to be used for the development of the RSA on the soil
surface at the study area. The measured PGA data from
seismometers were collected from Malaysia
Meteorological Service [15]. The comparisons were
done by looking to the percentage different DP%
between both values using Eq. (1) [1].

100
PGA Measured
PGA Measured - PGA Predicted
%
|
.
|
\
|
= DP

(1)

Where DP% is percentage of different and
measured, PGA is the peak ground acceleration
recorded by the seismometers in positive value.
The attenuation relations for each subduction and
fault zone that produced the lowest percentage of DP%
with positive values were then chosen as the
appropriate or suitable attenuation relationship used
for the study area.

1504
2.4 Design Response Spectra
The design response spectra is typically the starting
point of most codified seismic design and assessment
procedures and is predominantly used to prescribe the
applied inertia forces induced by earthquake ground
motions [16]. Design response spectra representing the
response spectrum of acceleration, in which the RSA
is used as an input [5]. In this study, NERA program
was used to develop RSA (response spectra of
acceleration) for the specific location in study area
where the results from soil data analysis are used as an
input parameter in generated the RSA before the
design response spectra based on available code can be
developed. Besides that, the determination of AF
(amplification factor) and site classification also
involved in development of design response spectra in
study area. In this study, the results from soil data
analysis such as value of unit weight was taken from
Coduto [17] and shear wave velocity was determine
based on empirical equations from Ohta and Goto [18],
as shown in Eq. (2).
V
S
= 85.3 N
0.341
(2)
where, V
S
is shear wave velocity and N is value of soil
penetration.
In order to determine the site classification (based
on NEHRP 2000 Provision/UBC 97 site classification)
and amplification factor in study area, Eqs. (3) and (4)
as follows are used, respectively:
:
s
=
d
i
n
i=1
d
i
si
n
i=1
(3)
J
n
=1
is equal to Su m
where, v
si
is the shear wave velocity in m/s and d
i
is the
thickness of any layer between 0 to 30 m from surface.
Amplication factoi =
PSA
PuA
(4)
where, PSA is peak surface acceleration and PGA is
peak ground of acceleration of the bedrock.
Uniform Building Codes 1997 or known as (UBC
97) are used in this study in order to determine the
response spectra in study area.
3. Results
Results obtained throughout this research are
presented into two phases accordingly which are Phase
1: PGA (peak ground acceleration) and Phase 2:
Design response spectra
3.1 Phase 1: PGA
According to Table 1, it can be seen that the PGA
values for subduction zone ranged from 0.0172 g to
0.0695 g while for fault zone ranged from 0.0187 g to
0.0314 g. Petersen et al. [8] (for subduction zone) and
Megawati et al. [10] (for fault zone) attenuation
relationships were taken as the most appropriate
attenuation model for this study based on the
capability of the equations to predict the PGA values
which were closest to the measured PGA recorded by
Malaysian Meteorological Service. Therefore, PGA
values of 0.0172 g and 0.0187 g were selected for
subduction and fault zone respectively in this study.
The maximum value of PGA (0.0187 g) is taken as
proposed PGA value that will be used for further
analysis in phase 2 of this study.
3.2 Phase 2: Design Response Spectra
Results obtained in this phase are summarized in
table form as follows:
It is found that most of study locations soil classified
into S
D
(Table 2). By referring to Table 3, Bandar
petaling Jaya was found to be in critical condition with
maximum AF 3.74 for stiff soil and Mutiara
Damansara with AF of 2.67 for S
D
. The peak RSA
(response spectrum acceleration) of 0.30 g (S
D
) for

Table 1 Summary of PGA values for subduction and fault
zone based on various attenuation relationship.
Zone Attenuation PGA (g)
Subduction
Megawati et al. [2] 0.0695
Lam et al. [7] 0.0564
Adnan et al. [6] 0.0259
Petersen et al. [8] 0.0172
Fault
Megawati et al. [10] 0.0187
Lam et al. [7] 0.0558
Campbell [9] 0.0314

1505
Table 2 Summary of data site classification for study
location.
Locations Borelog
Site classification
S
A
S
B
S
C
S
D
S
E
Bandar Puteri
Puchong
10 - - - 10 -
Mutiara Damansara 10 - - 1 9 -
Bandar Petaling Jaya 8 - - 1 7 -

Table 3 Summary of AF results.
Locations
AF
S
C
S
D
Bandar Puteri Puchong - 2.67
Mutiara Damansara 2.67 3.74
Bandar Petaling Jaya 2.14 3.74

Table 4 Summary of proposed design response spectra for
each site.
Locations
S
C
S
D

Peak of
RSA (g)

Period
range (s)

Peak of
RSA (g)

Period
range (s)
Bandar Puteri
Puchong
- - 0.18 0.09-0.47
Mutiara
Damansara
0.25 0.07-0.35 0.23 0.12-0.59
Bandar
Petaling Jaya
0.23 0.08-0.38 0.30 0.10-0.52

Bandar Petaling Jaya has been found as the highest
value of RSA with period range of 0.10 s to 0.52 s and
peak RSA of 0.25 g (S
C
) at period range 0.07 to 0.35
seconds for Mutiara Damansara (Table 4). The long
period range about 0.47 s with peak RSA 0.23 g (S
D
)
show by Mutiara Damansara.
4. Conclusions
Based on the development of design response
spectra based on various attenuation relationships at
specific location in this study, it can be concluded S
D
has a long range of period at peak spectral acceleration
compared to S
C.
where out of three locations, Mutiara
Damansara has a longer periods at 0.12 s to 0.59 s for
peak spectral acceleration of 0.32 g which is classified
as stiff soil. All these values are important in analyze
the structural design of building that has a period of
vibration at this range of period which is particularly
for building at all three locations in this study namely,
Bandar Puteri Puchong, Mutiara Damansara and
Bandar Petaling Jaya.
Acknowledgments
The authors would like to express their million
thanks to School of Engineering University Sains
Malaysia for giving full cooperation, support and
encouragement throughout the project.
References
[1] A.F. Kamarudin, Site specific response spectrum
acceleration due to Sumatran earthquake: Science and
technology complex, M.Sc. Thesis, Universiti Teknologi
Mara, Malaysia, 2006.
[2] K. Megawati, T.C. Pan, K. Koketsu, Response spectral
attenuation relationships for Sumatra-subduction
earthquakes and the seismic hazard implications to
Singapore and Kuala Lumpur, Soil Dynamics and
Earthquake Engineering 25 (2004) 11-25.
[3] T.C. Pan, Estimation of peak ground acceleration of the
Malay Peninsula due to Sumatra earthquake, Bulletin of
Seismological Society of America 92 (3) (1998)
1082-1094.
[4] MMS, Seismic Activity in Malaysia, Malaysia
Meteorological Service, Kuala Lumpur, 2005.
[5] M.N. Fadzli, Development of design response spectra for
Penang Island, M.Sc. Thesis, Universiti Sains Malaysia,
2007.
[6] A. Adnan, H. Hendriyawan, A. Marto, M. Irsyam,
Selection and development of appropriate attenuation
relationship for Peninsular Malaysia, in: Malaysian
Science and Technology Congress (MSTC), 18/20 April
2005, Cititel Hotel, Kuala Lumpur, Midvalley,
2004.
[7] N.T.K. Lam, A.M. Chandler, J.L. Wilson, G.L.
Hutchinson, Response spectrum prediction for potential
near-field and far-field earthquake affecting Hong Kong:
Rock sites, Soil Dynamics and Earthquake Engineering 22
(2002) 47-72.
[8] M.D. Petersen, J. Dewey, S. Hartzell, C. Mueller, S.
Harmsen, A.D. Frankel, K. Rukstales, Probabilistic
seismic hazard analysis for Sumatra, Indonesia and across
the Southern Malaysian Peninsula, Tectonophysics 390
(2004) 141-158.
[9] K.W. Campbell, Prediction of Strong Ground Motion
Using the Hybrid Empirical Method: Example Application
to Eastern North America, Bulletin of the Seismological
Society of America, USA, 2002.
[10] K. Megawati, T.C. Pan, K. Koketsu, Response spectral
attenuation relationships for Singapore and the Malay
Peninsula due to distant Sumatra-fault earthquakes,

1506
Earthquake Engineering and Structural Dynamics 32
(2003) 2241-2265.
[11] J. Sun, T.C. Pan, Seismic characteristics of Sumatra and
its relevance to Peninsular Malaysia and Singapore,
Journal of Southeast Asian Earth Sciences 12 (1-2) (1995)
105-111.
[12] K. Sieh, D. Natawidjaja, The seismic threat posed by
faults in Sumatra to Singapore and its neighbors, in:
Proceedings of the 8th

East-Asia Pacific Conference on
Structural Engineering and Construction, Singapore, Dec.
5-7, 2001.
[13] Active Tectonics and Seismologic Summary of the
Sumatra Plate Boundary, 2006, www.gps.caltech.edu
[14] V.Y. Sokolov, Site & region-specific response spectra: A
probabilistic approach, Soil Dynamics and Earthquake
Engineering 20 (2000) 273-281.
[15] Seismic Database Record 2006, Malaysian Meteorological
Service, Geophysics and Tsunami Division, Petaling Jaya,
Malaysia, 2006.
[16] A. Chandler, N. Lam, J. Wilson, G. Hutchinson, Review
of modern concepts in the engineering interpretation of
earthquake response spectra, in: Proceedings of the
Institution of Civil Engineers: Structures and Building,
Feb. 2001, p. 146.
[17] D.P. Coduto, Geothecnical Engineering Principle and
Practices, Prentice Hall, United State of America, 1999, p.
99.
[18] Y. Ohta, N. Goto, Empirical shear wave velocity
equations in terms of characteristic soil indexes,
Earthquake Engineering and Structural Dynamics 6 (1978)
167-187.


Structural Retrofitting of Old Illyria Hotel
Feti Selmani
1
and Musa Stavileci
2
1. PAN Engineering, Sofali 10000, Kosovo
2. Faculty of Civil Engineering and Architecture, University Hasan Prishtina, Sofali 1000, Kosovo

Abstract: Illyria hotel (formerly Bozhur) was built during the sixties in the heart of the actual Kosovos capital Prishtina, according to
former old Yugoslav standards in a Modernist architectural style. It represents a massive structural system with brick walls up to 54 cm
thick and avramenko type reinforced concrete floors. The investors aim was to add another two floors on the top of the existing ones
and to build two level underground parking floors, a health spa centre, whilst at the vicinity of the existing building (the northern side)
to erect a new 17 story-high brand new hotel and administration building. The retrofitting of the structure as well as construction of the
new structure has been done in full accordance with the new structural Eurocodes recommendations. 3D FEM (finite element
method) modeling was used for the analysis and design, using ETABS v 9.5 nonlinear and ARSAP 2010 (Autodesk Robot Structural
Analysis Professional 2010). Response spectrum design according to EC 8,3.2.2.4 has been used for seismic analysis and design with a
reference peak ground acceleration on type A ground of A
gR
=0.25 g.

Key words: Structural retrofitting, reinforced concrete shear walls, foundation strengthening, column jacketing.

1. Introduction

The old Illyria hotel (formerly Bozhur) is a
landmark building situated in the very centre of
Kosovos capital Prishtina (Fig. 1), designed according
to the modernist architectural style. The building was
designed and built during the late sixties according to
the old YugoslavJ US standards, and consisted of two
underground floors, a ground floor and another five
floors above. The architectural design represents a set
of four (4) building blocks: blocks A and B (the old
hotel), block C (the new hotel and administrative
building) and block D (the underground parking lot, a
spa centre and a conference hall on the first floor). The
owners wish was the erection of yet two other floors
on top of the existing ones (blocks A and B). This task
was the most challenging one, since the old structure
had already been in a poor state. After having
thoroughly examined different possibilities, three
possible solutions arose, namely:
(1) the demolition of the old building and the

Corresponding author: Feti Selmani, M.Sc., research fields:
structural engineering, passive control and bio-structures.
E-mail: panengineering@gmail.com.
erection of a new one;
(2) structural retrofitting of the existing structure and
the erection of the two stories on top of the existing
ones.
The second possibility was chosen in discussions
with the owner and the architect. In these
circumstances, three possibilities were analyzed:
(1) retrofitting of the old structure with CFRP
(carbon fibresreinforced polymers);
(2) steel X bracings;
(3) conventional structural elements jacketing.
The later approach was chosen mainly because of
economic as well as knowledge reasons, since
reinforced concrete is a well known material here in
Kosovo, whereas the two first possibilities require an
extended schedule and a different level knowledge.
The challenge was to find an original solution to the
problem by designing a completely new reinforced
concrete structure extending from under the existing
foundations to the top of the newly added floors (level
+ 25.89 m). Having into consideration the poor
physical state and mechanical characteristics of the
existing concrete (MB 15.20) as well as clay blocks
DAVID PUBLISHING
D

1508

Fig. 1 The old Illyria hotel (formerly hotel Bozhur)photo
taken around seventiesa modernist style icon in the heart
of Kosovos capital Prishtina.

and clayed bricks, the newly erected structure had been
analyzed and built fully independent (except the
foundations) from the existing one and serves
exclusively to support (both vertically and laterally) the
two new top floors. Laterally, the structure is stabilized
through new shear walls which run in two principal
directions and only a monolithic connection with the
middle row of the existing columns situated on the fifth
floor below has been made.
2. Geotechnical Conditions
After careful examinations of the site, seven
boreholes have been drilled: three at the southern side
and four at the northern side of the future building
blocks, whereas seven points with static penetration
and two points with geophysical-geological
measurement have also been taken in order to fully
assess the underground soil conditions for the
construction site [1]. The first four meters (borehole B2)
consisted of silt and sandy clay with medium to high
plasticity (in hard to very hard consistency) of brown
color (Fig. 2a), whilst at a depth of approximately 8 m
(the level of the underground water table) a mixture of
well-granulated gravel and sand was detected [2]. A 60
cm thick foundation slab has been designed for the
block D (wellness centresee Fig. 3) at this level
(SL_001, axes A-L/8-5). A soldier bored pile wall
with a pile diameter of 600 mm, distant 1,200 mm

(a)

(b)
Fig. 2 (a) photo showing the soil samples; (b) soldier bored
pile wall with a pile diameter of 600 mm for the purpose of
the ground support provision.

Fig. 3 Layout of Illyria complex building blocks: A and
Bexisting building (old hotel); Cthe new 17-stories hotel
and Dthe spa centre.

1509
axially [3], was used for the provision of the ground
support on permanent basis (Fig. 2b).
The bearing capacity of the soil is given as a function
of the boreholes position, and varies from 250 300
kPa with a safety coefficient of 3 [1]. A subgrade
modulus (k
z
=
0
/s) of k
z
=10,000 20,000 kN/m
3
had
been taken for analysis purposes. Thus, for analysis and
design of block C (the new 17 storey hotel) a k
z
=
10,000 kN/m
3
value of the sub-grade modulus was
considered, whereas the soil stress resulting from the
analysis is
soil
= 382 kN/m
3
. A layer of well
granulated and well compacted gravel 80 cm thick (in
layers of no more than approximately 20 cm thick) had
been laid under the foundation slab prior to the laying
of reinforcement and concrete.
3. Structural Analysis and Design
Structurally, Illyria complex is composed of four
independent dynamic units: block A and B (old hotel),
C (new hotel and administrative building) and D (spa
centre) as shown in Fig. 4.
3.1 Materials
Neither reinforcing details nor other drawings could
be found after the first author has searched in the citys
archives. Samples of existing reinforced concrete were
taken at various places on the existing structural
elements and after the examinations a mean value of
C16/20 concrete class could be established at the best
possible case.
A C25/30 (EC2-1-1, cl. 3.1) concrete class was
adopted for all reinforced concrete structural elements,
such as beams and structural slabs, whereas a C30/37
concrete class was adopted for shear walls as well as
the vertical elements such as reinforced concrete
structural columns (columns jacketing also) for
building blocks A, B, C and D. Reinforcing steel of
high ductility class B500B (EC2-1-1, cl. 3.2) had been
used for analysis and design [4].
Structural slabs for block C are made of reinforced
concrete, 25 cm thick (in accordance with EC 2,4.1.3
and EC 2,7.4.2, Table 7.4N), whilst for the blocks A
and BA prefabricated solution was adopted,
consisting of a 6 cm layer of concrete poured in situ on
the top of lightweight and fast forming system with
EPS (extruded polystyrene) filling prefabricated panels.
The main reason for choosing this system was to
contain structural dynamic mass as low as possible.
The partial factors for materials have been considered
from the respective parts of EC 1992 to EC 1998.
3.2 Actions
All actions for analysis and design were considered
in full accordance with the requirements of EC1
1991-1-1 to EC1 1991-1-7, while the combinations of
actions were considered according to EC 1990, cl.
6.4.3.2 6.4.3.4 for various design situations [5].
The imposed loads were taken into account in
accordance with Table 6.1 of EC1 1991, whilst the
values are taken from Table A.1.1 of EC 1990 [6].
Seismic action was considered in accordance with
the requirements of (EC 1998, cl. 3.2.1) with a
recommended value (from the part of a geotechnical
expert) of the reference peak ground acceleration on
ground type A a
gR
=0.3 g, where g =9.81 m/s
2
. Thus,
the design value of ground acceleration on type A
ground results as a =r
1
a
gR
=1.2 0.3 g, where: r
1
is
the importance factor and according to EC 1998, cl.
4.2.5, Table 4.3, importance class II has been assumed,
therefore the recommended value of r
1
=1.20. The
behavior factor q =3.5 was taken, as the structure was
designed as a DCM (ductility class medium) structure
in accordance with the recommendations of EC 1998,
cl. 5.4 [7]. The horizontal components of seismic
action are introduced as of Type 1 according to EC
1998, cl. 3.2.2.5, eq. 3.12 to 3.15 for a ground type B
(in accordance with the geotechnical report), whereas
the vertical component of the seismic action had not
been considered (see EC 1998, cl. 4.3.3.5.2).
The inertial effects of the seismic action were
considered as per expression (3.16 of EC 1998, cl.
3.2.4), whilst the combination of actions is in

1510
compliance
SSRS (squar
effect due
respectively
3.3 Analysis
State-of-th
2010, were u
Results, suc
displacemen
an outcome
(finite elem
structure w
freedom pe
rotational D
stories (level
structure (bl
forming sys
supplemente
order to incr
selection of
mainly on th
the floor is s
SL_700a) a
level is at -4
For dynam
slabs are mo
degrees of
displacemen

Fig. 4 Layo
Dthe spa ce

with EC 199
re root of the
to each ho

).
s and Design
he-art softwa
used for analy
ch as periods
nts were conf
there were sa
ment method)
was considere
er node (thre
DOFs). Struc
l + 17.19 2
locks A and B
stem for con
ed with add
rease the deck
f the above-m
he fact of dec
situated at a c
above the fo
4.06 m, see Fi
mic analysis
odeled as rigi
f freedom
nt degrees o
ut of Illyria co
entre.
S
8, cl. 4.3.3.5
e sum of squa
orizontal co
are ETABS nl
ysis and desig
s of vibration
fronted with e
atisfactory re
mathematic
ed, with its
ee displacem
ctural slabs f
25.89 m) on to
B) are of ligh
crete floors
ditional rebar
k resistance. T
mentioned sy
reasing the st
onsiderable h
oundations l
ig. 5).
purpose, str
id diaphragm
per story
of freedom a
omplex buildin
Structural Ret
(2), that is by
ares) of the ac
omponent (
l v.9 and ARS
gn of the struc
n, frequencie
each other an
sults. A 3D F
cal model of
six degrees
ments plus t
for the two
op of the exis
htweight and
type, with
r in-between
The reason for
ystem was ba
tructural mas
height (+20.7
evel (founda
ructural conc
ms with only t
(two horizo
and a rotati
ng blocks: A a
trofitting of O
y the
ction

SAP
cture.
es or
nd as
FEM
f the
s of
three
new
sting
d fast
EPS
n, in
r the
ased
s, as
9 m,
ation
crete
three
ontal
onal
deg
dow
nee
dec
on
add
resu
mod
Fig
stru
clea
mot
is
acc
stat
4. C
A
(DC
rein
(Fig
A
app
wou
floo
stre
a ty
colu
bas
and Bexistin
Old Illyria Hot
gree of freed
wnsizing of t
eded as we
ision-taking
a local or on
dition, for the
ults, what had
del of the stru
. 5 shows t
ucture of bu
arly the distin
tion (translat
an indicatio
ording to EC
ted by it).
Constructio
As the structu
CM), special
nforcement fo
g. 6) [7].
According to
proximately 2
uld have bee
ors on top
engthening be
ypical detail o
umn jacketin
ement level).
ng building (ol
tel
dom). The r
the enormou
ell as to
relatively to
n a global e
purpose of co
d been consid
ucture (for all
the first thr
uilding block
nction betwee
tions) and the
on [7] of t
C8, cl. 4.2.3
on and Det
ure was design
attention wa
or columns, b

o calculation
0% of the fou
en exceeded
of the exis
ecame necess
of foundation
ng through la
.
ld hotel); Ct
reason for t
s amount of
increase the
the stiffness
xtent of the
omparison of
dered is a 3D m
l building blo
ree mode sh
k C, where
en the two fir
e third one (t
the structura
.2 (although
tailing
ned of a med
as paid to the
beams and th
ns for the o
undations bea
by adding
sting ones,
ary. In Fig. 7
n strengthenin
attice metal s
the new 17-sto
this was the
f calculations
e speed of
deployment,
structure. In
f the obtained
mathematical
ocks A, B, C).
hapes of the
one can see
rst periods of
orsion). This
al regularity
h not directly
dium ductility
e detailing of
e shear walls
old building,
aring capacity
another two
therefore, a
7, one can see
ng as well as
structure (the

ories hotel and
e
s
f
,
n
d
l
.
e
e
f
s
y
y
y
f
s
,
y
o
a
e
s
e
d

1511

Fig. 5 Layouts of the 3D mathematical modelbuilding block C; the first three mode shapes of vibration: T
1transl
= 1.82 s,
T
2transl
= 1.51 s, T
3torsin
= 0.75 s (ARSAP 2010).

(a) (b)
Fig. 6 (a) The structural columns reinforcement for the critical regions (EC 1998, CL.5.4.3.2); (b) reinforcement for the
foundation slabbuilding block C.

Fig. 7 Left: column jacketing (at the basement levelblock A); Right: foundation strengthening.
(a) Formworks
(b) Reinforcement

1512

(a) (b)
Fig. 8 The existing reinforced concrete columns: (a) column jacketing (at the basement levelblock A); (b) foundation
strengthening.

(a) (b)
Fig. 9 (a) Shear wall and column jacketing tied together; (b) shear wall under construction, Block Btop floor.

Whilst the foundations of the blocks A and B were
strengthened with reinforcement bars, the existing
reinforced concrete columns were strengthened
(Figs. 8a and 8b) with an original metal lattice
confining structure as further shown in Figs. 7 and 8.
New shear walls are connected with existing r. c.
structure as shown in Fig. 9.
5. Conclusions
(1) What had initially been considered is the
structural retrofitting with fibre reinforced polymers
such as carbon fibres or x metal bracing. However, the
authors are satisfied that the targeted compliances were
achieved using conventional solutions such as
jacketing, shear walls, etc. with a lower priceusing
Existing
New
Typical section

1513
local materials, skills and resources;
(2) The principal challenge of preserving an old
landmark structure in the historic part of the city
through retrofitting its structure and putting it in-line
with the modern code requirements (Eurocodes)has
been successfully achieved and the client received the
targeted product;
(3) The first author have personally achieved and
gained professionally, whilst the reliability of the
structure has been increased in cases of extreme natural
hazards such are earthquakes under a very tight time
schedule and constraints;
(4) Nowadays, the name of the hotel is no longer
Illyria Hotel but it changed into Swiss Diamond Hotel
[8], and it is certainly the best hotel in the City of
Prishtina.
References
[1] EN 1997-1:2004/AC:2009, Eurocode 7: Geotechnical
DesignPart 1: General Rules, Comit Europen de
Normalisation (CEN), Bruxelles, 2009.
[2] EN 1991-1-1:2002/AC 2009 Eurocode 1: Actions on
StructuresPart 1-1: General ActionsDensities,
Self-weight, Imposed Loads for Buildings, Comit
Europen de Normalisation (CEN), Bruxelles, 2009.
[3] EN 1991-1-1:2003/AC 2009 Eurocode 1Actions on
StructuresPart 1-3 to 1-7, Comit Europen de
Normalisation (CEN), Bruxelles, 2009.
[4] Eurocode 2: Design of Concrete StructuresPart 1-1:
General Rules and Rules for Building, Comit Europen
de Normalisation (CEN), Bruxelles, 2004.
[5] EN 1998-1:2004/AC:2009, Eurocode 8: Design of
Structures for Earthquake ResistancePart 1: General
Rules, Seismic Actions and Rules for Buildings,
Comit Europen de Normalisation (CEN), Bruxelles,
2009.
[6] GEOINGGeomechanical Report, Complex Illyria,
Prishtin, Hristo Georgijevski responsible engineer,
Skopje, 2009.
[7] U. Smoltczyk, Geotechnical Engineering Handbook,
Volume 3, Ernst & Sohn, Berlin, 2003.
[8] Swiss Diamond Hotel Web Site,
www.swissdiamondhotelprishtina.com (accessed Feb. 12,
2013).


Challenges of Structure in Todays Architectural,
Economic and Social Context
J orge Bernabeu Larena
1
and Alejandro Bernabeu Larena
2

1. Faculty of Civil Engineering, Technical University of Madrid, Madrid 28004, Spain
2. Faculty of Architecture, Technical University of Madrid, Madrid 28004, Spain

Abstract: The recent establishment of a digital culture and society, together with current financial crisis and urgent energetic and
climatic needs, has radically changed the architectural scene from the optimism of some years ago to a situation of uncertainty and
huge social demands and challenges. In this context, it is suggested to rethink the role of structure in architecture, such as an enabler,
a guide and a catalyst. The purpose of this paper is to present the economic, cultural and social context in which architecture
develops nowadays. The method, to suggest a discussion on which role the structure may adopt in the architecture to come. The
achievement, to highlight its potential to face current requirements and challenges.

Key words: Structural engineering, design concept, social architecture, critics.

1. Introduction

The end of 20th century and the beginning of 21st
has been characterized in architecture by a strong
development of design, analysis and construction
technologies. This technical development has
generated an unprecedented situation of great freedom,
in which almost every form can be solved and erected.
Frank Gehrys Guggenheim museum in Bilbao or
Herzog & de Meurons Beijing Stadium are
paradigmatic examples of this architectural freedom
[1].
In this context, being the structural requirements
overcome, the purpose of structure and its relation to
architecture had to be reconsidered. And indeed, while
in earlier times, the development of architectural form
was directly related to the emergence of new
structural materials or systems, the new technical
control and formal freedom required a further
consideration on the relation between architecture and
structure, and on the potential of structure as a design

Corresponding author: J orge Bernabeu Larena, Dr., Eng.,
research fields: history of construction, public works heritage,
cultural landscape and structural engineering. E-mail:
jorge.bernabeu@upm.es.
tool. New structural design strategies had to be
developed, in order to consider structural behavior and
requirements not as annoying constraints, but as
design opportunities, within which develop
architectural form and composition [2].
However, the amazing development of digital
technologies and its strong establishment in society,
together with the urgent energetic and financial needs,
have radically changed the scene. From the optimistic,
and maybe irresponsible, architectural freedom of the
beginning of the twenty-first century, people have
now moved into a situation of architectural
uncertainty, in which the design parameters and
priorities of just some years ago are no longer valid,
and have to be revised, reordered or redefined.
And in this context, it is suggested once again to
rethink the role of structure in architecture, both as an
enabler, a guide or a catalyst. The final goal of this
paper is to present the economic, cultural and social
context in which architecture develops
nowadaysThe research procedure is based on a
discussion about the role the structure may adopt in
the architecture to come. What is new in this study is
the critic comparative analysis between different
DAVID PUBLISHING
D
Challenges of Structure in Todays Architectural, Economic and Social Context

1515
contemporary projects digital culture, society and
architecture.
The recent establishment of a digital culture is
radically changing the social and cultural context in
which architecture evolves, transforming its main
parameters and factors [3]. Among them, it is
interesting to highlight the following:
Redefinition of the relation to the physical world.
It entails the transformation of the concept of
materiality, in benefit to virtual reality, from spatial
recreation to cyber-sociality, and the relation or
opposition between real and virtual (Fig. 1);
Emphasis on individual preferences and choices,
altogether with the establishment of a globalized
culture and society. The widespread introduction of
globalized products and systems that may be
personally set and customized, reveals this duality
between standard and unique, global and personal,
collective and individual (Fig. 1);
Relevance of events. What happens, the content,
becomes more important than any physical thingthe
container. As an example, the impressive
establishment of the internet, with instant update of
contents following the last events and news, takes the
user into a vertiginous process that gives the feeling of
living in a permanent present.
Related to architecture, these factors entail the crisis
of traditional tectonic, and its relation and dependence
to physical materiality, claiming for the establishment
of a new tectonic, able to reconcile abstraction of
virtual society and culture with the physical
concretion that architecture requires [1].
Many projects have been presented in recent years
that can be considered in a broad sense as digital
architecture, showing architects interest and concern
on digital demands and motivations, and illustrating
also the context of architectural freedom (Fig. 2).
Among them, Zaha Hadid, together with his partner
Patrick Schumacher, has proposed an ambitious
theoretical argument or manifesto. It is based in
integrating in the development process of a project a
number of factors and complexities, related not only
to the project or to construction and architectural
techniques, but to all kind of economic, social and
cultural motivations and suggestions [4, 5]. This
parametricism proposal, that has been presented as
the new architectural style after modernism, proposes
an understanding of architecture as the result of a set
of different factors and parameters that, combined,
differentiated and correlated among them, defines a
malleable formal development. It identifies besides a
number of principles from which project design
development arises, and that are organized as negative
or taboos (rigid forms, repetitions, etc.) and positive
or dogmas (soft forms, differentiated and
interdependent systems, etc.). It finally raises the
question about structure, if it is just another parameter
to be considered in the equation, or if it can actively

Fig. 1 Digital social and cultural context, virtual reality and cyber-sociabilitysecondlife, duality between standard/global
and unique/personalipod.

1516

(a) (b)
Fig. 2 (a) Digital architecture, Soft Office, Warwickshire (UK), 2002. NOX; (b) Mercedes Benz Museum, Stuttgart
(Germany), 2006, UN Studio.

participate in the definition of the patterns that govern
and control the parametric development of the project.
It is maybe too early to evaluate the result and the
influence of these proposals in the architecture to
come. However, they still lack, in its practical
application, of the coherence and in depth analysis and
development that would be expected from their
theoretical considerations. Overly concerned by
formal and visual aspects, in accordance to digital
imagery of nowadays [6, 7], they avoid to face in
depth the complex challenges that digital architecture
proposes, as link and catalyst between virtual society
and the tectonic and material reality of built
environment.
Accordingly, the structure of these project lacks on
its side of new approaches or strategies, and refuse to
lead the process in depth, concentrating in adequately
resolve the forms proposed by the architects. The
complex geometries of Phaeno Science Centre, by
Zaha Hadid and Graz Music Theater, by UN Studio,
illustrate this situation. In both cases, the structure,
developed by two of the most relevant and innovative
engineers of nowadays (Hanif Kara and Cecil
Balmond, respectively), has been brilliantly solved, is
elegant, interesting and well adapted to architectural
forms and geometries [8, 9]. But it does not represent

Fig. 3 Phaeno Science Center, Wolfsburg (Germany), 2005. Zaha Hadid Architects/Adams Kara Taylor, Graz Music
Theater, Graz (Austria), 2008, UN Studio/Ove Arup & Partners.


1517

neither in both cases any relevant novelty on structural
systems or types, nor establish any link or approach to
the big questions that digital architecture raises
(Fig. 3).
2. Energetic, Climatic and Economic Needs,
The Architectural Response
Current financial crisis, altogether with increasingly
urgent energetic and climatic needs, questions the
final purpose of contemporary architecture and
construction, and strongly suggests decay of visual
leadership in favor of functionality, resources
optimization and cost control.
The concern for the environment and the need to
preserve it, reducing the emission of pollutants, as
well as decreasing energy and natural resources
consumption, is present in society for several years
now. There are therefore many architectural projects
concerned by these issues, especially in reducing
energy consumption, both for environmental reasons
and for decreasing building operating costs. Their
interest has focused in many cases in the faade, that
is becoming more and more sophisticated. This
sophistication of the faade responds both to the
incorporation of advanced materials and climate
control or energy uptake systems, and to visual and
aesthetic motivations, related to technological and
digital imagery (Fig. 4).
Likewise, environmental certifications have gone
from being practically unknown in its beginnings in
architecture, twenty years ago, to be highly demanded
nowadays, confirming the concern, or at least the
interest, of society towards climatic and energetic
sustainability. However, there are several issues that
may be questioned when considering environmental
certification [10]. First of all, it is important to
consider to what extent it is possible to measure the
sustainability of a building. Secondly, if
environmental architecture consists just in complying
with the different certification criteria, it risks to be
excessively neutral and banal. And finally, the high
demand of environmental certificates by big brands
and organizations, denotes their concern not so far in
the environmental needs, but in being considered
environmentally responsible or friendly (Fig. 4).
Considering so, certification is no longer a mean, but
an end in itself, loosing much of its true potential.
In addition, or as an alternative, to these approaches,
the following two proposals, completely different one
from each other, illustrate and raise several issues that
are key for considering sustainable architecture:
First, Glenn Murcutt projects, widely
acknowledged since the Pritzker prize in 2002,
illustrate a straightforward, subtle and responsible
approach to sustainable architecture, not based on
incorporating complex energy control systems, but in
working in harmony with nature [11, 12]. And indeed,
what Murcutt architecture proposes is a sensitive
approach to nature and to natural rhythms, the respect
to landscape scale, the prevention of natural resources,

(a) (b)
Fig. 4 (a) Media-TIC offices building. Barcelona (Spain), 2010, Cloud 9; (b) Coca-Cola Headquarters, Madrid (Spain),
2009, De Lapuerta and Asensio, The building has achieved a gold LEED certificate.

1518

and above all, the awareness of the responsibility of
architecture in the transformation of the world that
entails any construction [13, 14].
Secondly, Fosters project for Masdar city in Abu
Dhabi, a sustainable, zero-carbon, zero-waste and car
free city, exemplifies the model of smart cities
projects that are proliferating nowadays, from Europe
or the United States to Latin America or Asia [15].
Besides the technological and somehow futuristic
approach of the project, what is interesting to
highlight here is the idea of facing environmental
needs in a bigger scale than building construction,
aware of the fact that the garden city and urban sprawl
models prevent any sustainable program, and that new
city developments have therefore to be adopted
[16, 17]. Accordingly, Masdar city proposes a high
urban density model, with controlled buildings height,
by reducing street width and incorporating an efficient
underground public transport network (Fig. 5). As a
result of this model, the need to move in private
transport decreases, and therefore energy consumption
and pollution is reduced.
Moreover, worldwide financial crisis has radically
changed the optimistic and even unconcerned
panorama of just some years ago, passing to a
situation of economic restraint, and placing architects
in the spotlight [18]. Terms and conditions in which
projects are developed have indeed significantly
changed, affecting the way architects work, while the
role and responsibility of architecture towards society
is more evident than ever.
To this demanding situation, architecture has not
been able yet to decisively respond. It is aware that the
situation has changed, and that it can no longer
consider the same project parameters and principles
that some years ago, that living and working
conditions have to evolve. But it has not been able yet
to clearly formulate the new criteria that would guide
design process. It is probably premature therefore to
evaluate and to draw conclusions on how might
architecture responds concretely to the current
financial crisis.
However, if people attend to projects carried out on
developing countries or degraded neighborhoods, that
share some of the needs arising from current crisis
situation, especially in terms of economic restraint and
efficient use of resources, them might extract some
findings that help to frame the situation.
In this sense, projects by Dibdo Francis Ker in
Burkina Faso and Alejandro Aravena in Chile are
especially significant.
First, Francis Ker projects have gained
international recognition for its successful
combination of modern construction techniques and
local traditional wisdom and materials, for the
minimum resources employed, and for the architects
complete implication in the project [19].
Indeed, for Gandos primary school project, built in
2004 and subsequently extended in 2008, Francis
Ker conceived the project, raised the necessary funds

(a) (b)
Fig. 5 (a) Arthur & Yvonne Boyd Art Center, Riversdale (Australia), 1998, Glenn Murcutt; (b) Masdar City, Abu Dhabi
(United Arab Emirates), 2007, Foster & Partners.

1519

to be carried out, designed it, organized the
construction involving his neighbors and future users
of the school, that worked as bricklayers and
construction laborers, and supervised the works
(Fig. 6). The result is a compromised, responsible and
sustainable architecture, that satisfy people needs, that
efficiently mitigates the stifling climate of the area,
and that is configured and realized respectfully of
local traditions and open to modern techniques.
For his part, Alejandro Aravenas work with
Elemental focuses in the design of innovative models
for public housing in Chile [20, 21]. The goal is to
respond to two main concerns: the awareness of
housing as an urgent social issue, and the promotion
of a sustainable development of the city. To respond
simultaneously to these two issues. Elementals
proposal is directly related to the idea of returning to
the dense city, that has proved to be more sustainable,
in terms of transportation, energy consumption and
pollution generation, as has been discussed before,
and that will bring inhabitants of shantytowns back
towards the city center, favoring their integration and
participation in the city life. Besides, the economic
restraint of these projects made that there was no
budget enough for building reasonable houses, in
terms of quality of construction and surface.
Their inventive and social response consisted then
in developing an open housing typology, actually built
with the essentials for a minimum unit, and the void
of a vacant space that can be filled afterwards by the
inhabitants. The first realized project with this
typology was built in Iquique in 2004, where the
layout took the form of a collective row housing with
an upper crenellated silhouette. The idea was then that
the residents would build in the remaining void of the
upper level, as their means allow. And actually, just
one year after delivery most of the voids were filled in,
completing the project (Fig. 7).
Several issues are important to highlight from these
projects, that although not directly extrapolable to the
general architectural financial crisis situation, raise
some key elements for reflection and discussion, and

Fig. 6 Primary school extension, Gando (Burkina Faso), 2008, Dibdo Francis Ker.

Fig. 7 Resorption of a shantytown for 100 families, Iguique (Chile), 2004, Elemental, Alejandro Aravena.


1520

suggest possibilities for current development.
First of all, it is crucial in both projects the intense
participation and involvement of future users and
inhabitants in the project development and during the
construction process. The inhabitant is not just the
final user of the project, but has an active role in its
realization.
Secondly, the concept of beauty is not considered
per se, but related to functionality and to the response
to the needs. Beauty is in any case a consequence of
the project, not a reason or a motivation factor of it.
Finally, application of sustainability criteria, use
and optimization of local resources, and economic
restraint, are not motivated in these projects by
ideological reasons or to defend an idea, but just for
practical matters, because it is the most logical and the
only thing that can be done. It is not so much an
option, but the natural solution to a necessity.
3. The Challenges of Structure
The irreversible establishment of a digital culture
and society faces architecture and engineering to an
unprecedented challenge that entails the redefinition
of their basis: the concepts of tectonic and materiality.
The goal is to relate them to digital and virtual
parameters, changing somehow their present
principles and characteristics.
Besides, current situation faces architecture and
engineering to important ecological and economic
challenges, to which they have to respond with
responsibility and sensibility, conscious of its main
social goal of satisfying people needs. This ethic of
construction entails, among others, optimization of
local resources, reduction of energy consumption, and
consideration both of modern and traditional building
techniques, in order to satisfy functional requirements
with a minimum construction and maintenance cost.
They finally suggest to rethink the concepts of beauty
and aesthetics in architecture, and its dependence to
purely visual aspects.
Architecture has understood the magnitude of these
challenges, although it has not found yet the way to
definitely face them. In turn, structure has not taken
these challenges as their own, but adopted a secondary
role. Structure is responding to architectural demands,
in some cases accurately and with brilliant results, but
without taking the initiative. And nevertheless, some
of the suggestions are closely related to structure and
engineering concepts and classical principles: to the
way they relate to nature and integrate into the
landscape, to the consideration of beauty in relation to
functionality, or to the idea of optimization and
minimization of the structure.
And if the authors consider simultaneously the
projects of the Phaeno Science Center and the Primary
School in Gando (Fig. 8), it arises that there is much
work to be done to reconcile digital and virtual
aspirations with sustainable and ecological demands,
and that the answer to both concerns may be found by
looking at the matter as a whole, that the potentiality
and powerful possibilities of the first may be a key
element to accurately respond to the second.

Fig. 8 Phaeno Science Center. Wolfsburg (Germany), 2005, Zaha Hadid/Adams Kara Taylor, Primary school extension,
Gando (Burkina Faso), 2008, Dibdo Francis Kr.

1521

4. Conclusions
From these considerations, far from trying to
establish a global conclusion, the authors propose an
open discussion on the role of structure in
contemporary architecture and its potential in helping
it to successfully face current economic, social and
cultural requirements:
Is structure able to establish a new tectonic, relating
physical reality of construction to virtual culture?
Should structure step back, in benefit of services
and climate systems, conscious of current energetic
urgent needs?
Should structure act as an enabler to architectural
freedom, or as a guideor even a restrainerto it?
What role may structure play in the redefinition of
beauty and aesthetics in architecture, faced to current
situation?
And finally, is structure able to act as a catalyst,
helping to reconcile digital culture suggestions with
economic, social and environmental current
requirements?
The challenge that current situation proposes both
to architecture and structure is huge. It is about
redefining the way people inhabit, how they occupy
the world and use natural resources, and how they
relate to each other and to nature. Let us face it with
decision, coherence and responsibility.
References
[1] D. Azagra, A. Bernabeu, Free shape structures, Informes
de la Construccin (Construction Reports) 64 (526) (2012)
133-142. (in Spanish)
[2] A. Bernabeu, What we talk about when we talk about
structures, the function of structure in architecture, in: P.
Cassinello, S. Huerta, J.M. de Prada Poole, R. Snchez
Lampreave (Eds.), Geometry and Proportion in Structural
Design, Essays in Ricardo Arocas Honour, Lampreave,
2010. (in Spanish)
[3] A. Picon, Digital Culture in Architecture, Birckhuser,
Basel, 2010.
[4] P. Schumacher, Parametricism as styleParametricist
manifesto, in: The 11th Architecture Biennale, Venice, 2008.
[5] P. Schumacher, Let the style war begin, Architects
Journal 16 (2010) 41-45.
[6] J. Steele, Architecture and Digital Revolution, Gustavo
Gili editions, Barcelona, 2001. (in Spanish)
[7] J. Krauel, Digital Architecture, Innovation and Design,
Links, Barcelona, 2010. (in Spanish)
[8] H. Kara, Design Engineering AKT Adams Kara Taylor,
Actar, Barcelona, 2008.
[9] UN Studio in Motion, UN Studio, Phoenix Publishing
and Media Group, Tianjin Ifengspace Culture & Media
Co., 2012.
[10] E. Prieto, To think, to count, to measure, key aspects in
environmental certification, Arquitectura Viva (Live
Architecture) 130 (2010) 112. (in Spanish)
[11] P. Drew, Leaves of Iron, Glenn Murcutt: Pioneer of an
Australian Architectural Form, The Law Book Company,
Sydney, 1985.
[12] E. Croquis, Glenn Murcutt, 1980-2012, Plumas de
Metal = Feathers of Metal, El Croquis, 2012, pp. 163-164.
[13] L. Moix, G. Murcutt, Architecture: More for less, in: L.
Fernndez-Galiano (Ed.), Fundacin Arquitectura y
Sociedad (Architecture and Society Foundation)/
Arquitectura Viva (Live Architecture), Madrid, 2010, pp.
[14] F. Fromonot, The success of an anti-star, Glen Murcutt,
Pritzker Prize, Arquitectura Viva (Live Architecture) 83
(2002) 73-75. (in Spanish)
[15] S. Behling, Masdar, city of the future, Arquitectura Viva
(Live Architecture) 130 (2010) 28-31. (in Spanish)
[16] L. Fernndez-Galiano, The gulf and us, Dubai versus
Masdar: Lessons from the crisis, Arquitectura Viva 130
(2010) 66-67. (in Spanish)
[17] L. Fernndez-Galiano, Architecture and life, art in
mutation, in: The Conference at the Royal Academy of
Fine Arts of San Fernando, Madrid, 2012. (in Spanish)
[18] J.L. Mateo, After Crisis, Lars Mller Publishers, Baden,
Switzerland, 2011.
[19] L. Moix, D.F. Kr, Architecture: More for less, in: L.
Fernndez-Galiano, (Ed.), Fundacin Arquitectura y
Sociedad (Architecture and Society Foundation)/
Arquitectura Viva (Live Architecture), Madrid, 2010, pp.
[20] M.H. Contal, J. Revedin, Sustainable Design: Towards a
New Ethic in Architecture and Town Planning,
Birkhuser, Berlin, 2009.
[21] A. Aravena, Elemental: Incremental Housing and
Participatory Design Manual, Hatje-Cantz, 2012. (in
Spanish)


Simulation of Corrosion Crack in Pretensioned
Prestressed Concrete Members
Agus Maryoto and Takumi Shimomura
Department of Civil and Environmental Engineering, Nagaoka University of Technology, Niigata 940-2188, Japan

Abstract: Corrosion of PC tendon embedded in concrete deteriorates performance of PC (prestressed concrete) members. Corrosion
crack observed in pretensioned PC members is different from that in reinforced concrete and post tensioned PC members. The
purpose of this study is to clarify mechanisms of corrosion crack in pretensioned PC members. Contribution of compressive stress
induced by prestressing force and arrangement of PC tendons on corrosion crack in pretensioned PC members is investigated through
experiment and numerical analysis. Three dimensional finite element analyses were carried out to simulate corrosion crack in PC
specimen. Based on the analytical and experimental results, the influencing factor of corrosion crack in pretensioned PC members is
discussed. It was concluded that the contribution of compressive stress in concrete by prestressing force on corrosion crack is not
significant and that the dominant reason for wider corrosion cracks in pretensioned PC members is the arrangement of PC tendons
with small intervals.

Key words: Corrosion crack, compressive stress, corrosive product, PC tendon, pretensioned PC member.

1. Introduction
It is well known that corrosion of steel reinforcement
embedded in concrete causes concrete crack along the
reinforcement in RC (reinforced concrete) and PC
(prestressed concrete) structures. So far, most studies
on corrosion crack have focused on RC structures with
steel reinforcement. Recently, some studies on
corrosion crack in post tensioned PC structures with
steel sheath have been reported by Mutsuyoshi et al. [1]
and Miyagawa et al. [2]. Compared with the cases of
RC and post-tensioned PC, corrosion crack in
pretensioned PC structures have not been well clarified.
Fig. 1 shows pretensioned PC beams of Urokazaki
Bridge which had been in service for 35 years in the
coastal area in Niigata, Japan [3]. A number of
corrosion crack were observed on the surface of
concrete. The crack width reached 11.5 mm at most,
which is around three times of diameter of PC tendon
used in this member. The reason of such wide corrosion
cracks in pretensioned PC members is not clear.

Corresponding author: Agus Maryoto, M.T., Ph.D. student,
research field: technology of concrete. E-mail:
agus_maryoto1971@yahoo.co.id.
Therefore, the authors try to clarify it in this study. The
authors consider two possibilities as the reason of
wider corrosion crack in pretensioned PC members:
effect of compressive stress due to prestressing force
and arrangement of PC tendons with small intervals.
Compressive stress induced in concrete by
prestressing force can reduce or eliminate tensile stress
in concrete in PC members under service load. As a
consequence of compression stress in longitudinal
direction, concrete is tensioned to lateral direction due
to Poissons effect. This might accelerate corrosion
crack along the longitudinal direction. Study on
corrosion of PC tendon in concrete by chloride attack
was performed by Li et al. [4]. It was shown that the
magnitude of prestressing force affects the corrosion
rate due to mechanical chemistry effect.
On the other hand, pretensioned PC member
contains multi layers of PC tendons in both horizontal
and vertical direction. Contribution of the second or
higher layers of PC tendons to propagation of corrosion
crack might not be neglected. The effect of spacing
between reinforcements on propagation of corrosion
crack and crack width on surface of RC member was
DAVID PUBLISHING
D

investigated
the width of
and the leng
structures, in
corrosion of
concrete co
crack in pr
unique beha
PC and RC a
This stud
stress and a
crack in pre
analysis and
2. Investig
on Corros
2.1 Experim
2.1.1 Spec
Fig. 2 sho
compressive
specimens w
Table 1 sho

(
(
Fig. 1 Exam
Simulat
d by Tran et a
f corrosion cr
gth of rebars.
nvestigation b
f PC sheath
over. Howeve
retensioned P
avior and diff
as shown in F
dy investigate
arrangement
tensioned PC
d laboratory te
gation of t
sion Crack
mental Investig
cimen
ows the specim
e stress on c
which are PC-
ows the com
(a) Existing of
(c) Beam of Ur
mple of corrosio
tion of Corros
l. [5]. The res
rack is affecte
. In case of p
by Minh et al
causes crac
er, propagati
PC structure
fer from that
Fig. 1.
es the effect
of PC tendo
C members th
est.
he Effect o
gation
men to investi
corrosion cra
-0, PC-2 and P
mpressive str
Urokazaki Br
rokazaki Bridg
on crack in pre
sion Crack in
sults showed
ed by the spa
posttensioned
l. [6] showed
ck on surface
ion of corro
es might hav
of posttensio
t of compres
ons on corro
hrough numer
of Prestres
igate the effe
ack width. T
PC-4, were te
resses applied
ridge
ge after remov
etensioned PC
n Pretensione
that
acing
d PC
that
e of
osion
ve a
oned
ssive
osion
rical
ssed
ct of
Three
ested.
d in
con
T
shea
indu
of t
cast
mm
wei
mat
test
100
2
E
rust
wat
Fig
test
corr
Dur
wid
C
is c
ved from existin
C girder of Uro
ed Prestresse
ncrete.
The PVC (po
ath of post te
uce prestressin
the concrete
ting fresh con
m of each be
ight. Specime
ttress for 10 d
ed by the cyli
0 mm and 200
2.1.2 Electroly
Electrolytic co
t in short tim
ter containing
. 3 shows a s
t. Direct elec
rosion proces
ring the test, a
dth were recor
Crack on the b
considered.

ng
kazaki Bridge
ed Concrete M
lyvinyl chlor
ensioned PC b
ng force from
mixture is sh
ncrete into the
am was weig
ens were cure
days. Compre
indrical specim
0 mm height w
ytic Corrosio
orrosion test w
me. Specimen
g 3% sodium
schematic of
ctric current
ss. PC tendon
accumulative
rded.
bottom surfac
Occurrence
(b)
(d)
.
C
c
Members
ride) pipe was
beam to protec
m corrosion. T
hown in Tabl
e form, the PC
ghed to obta
ed being cove
essive strength
men whose di
was 30 MPa.
on Test
was conducte
ns were imm
chloride as a
the electroly
was applied
ns 6 mm we
e electric curre
ce of the spec
of cracks w
Cross section
) Front view
Corrosion
crack with
11.5 mm
width
1523
s utilized as a
ct steel bar to
he proportion
le 2. Prior to
C tendon 6
ain the initial
ered with wet
h of concrete
imension is

ed to produce
mersed in the
an electrolyte.
ytic corrosion
d to promote
ere corroded.
ent and crack
imen (Fig. 3)
was checked

3
a
o
n
o
6
l
t
e

e
e
.
n
e
.
k
)
d
Simulation of Corrosion Crack in Pretensioned Prestressed Concrete Members

1524

Table 1 Compressive stress.
Name of specimen Compressive stress (N/mm
2
)
PC-0 0
PC-2 2.33
PC-4 4.73

Fig. 2 Specimen to investigate the effect of compressive
stress.

Table 2 Proportion of concrete mixture.
Material Unit (kg/m
3
)
Water 154
Cement, high early strength type 390
Sand 622
Gravel 1,189
Additive, water reducing agent (ltr/m
3
) 3.9
Gmax, maximum size of gravel (mm) 25
w/c, water/cement weight ratio (%) 39.5
s/a, sand/total aggregate volume ratio (%) 35.5

Fig. 3 Schematic of electrolytic corrosion test.

every day. Occurrence of the first crack was judged
when crack width reached 0.01 mm. Thereafter, crack
width was measured using manual scale at 3 points: in
the middle of the beam, 100 mm to left side and right
side from the midpoint. Average of those 3 crack
widths was employed in the discussion.
Electric current induced in corrosion test was
recorded by the data logger every 30 min.
Accumulative current (IT) was calculated at the end
of corrosion test. Amount of corroded PC tendon can
be estimated using equation below based on Faradays
law.
= IT (1)
where, W is mass loss of PC tendon (gr), is
coefficient of electrolytic corrosion (gr/A.hr), I =
current (A) and T is time (hr). Coefficient of
electrolytic corrosion is 1.0744 gr/Ahr according to
= M / (z F), where M is the atomic mass of the metal
(Fe = 56 gr), z is the ionic valence of the metal (Fe = 2)
and F is Faradays constant (96,485 As). After the
corrosion test was completed, corroded PC tendons
were taken out from concrete. Rust was then removed
from PC tendons using a wire brush and then immersed
in 10% ammonium citrate solution in one day. Mass
loss of the PC tendon during the corrosion test was
calculated from the accumulative current at the time by
interpolation.
2.1.3 Test Results and Discussions
Fig. 4 shows experimentally obtained relationship
between corrosion of PC tendon and crack width of the
specimens under compressive stress. It is found in Fig.
4 that compressive stress contributes to the initiation
and opening of crack. Occurrence of crack in
prestressed concrete is a little faster than in specimen
without compressive stress. This phenomenon can be
led by Poissons effect, concrete beam is
under compression in the longitudinal direction due to

Fig. 4 Experimental result of crack width.
100
50 50
150
114
18
18
Concrete
900
50 50 800
PCTendon 6 for corrosion
Steel bar 11 for pre-stress
Steel plate
Bolt
PVC pipe 16
a. Cross section b. Front View
+
+
3%NaCl solution
copperplate
packer
PCTendon6
PCTendon6
concrete
Bottom surface
Topsurface
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 2 4 6 8 10 12
C
r
a
c
k

w
i
d
t
h

(
m
m
)
Corroded PC tendon ( x 10
-4
g/mm
2
)
Experimental, PC-0
Experimental, PC-2
Experimental, PC-4
(a) Cross section (b) Front view

1525
pre-stress and simultaneously under tension in the
lateral direction. However, the quantitative
contribution of compressive stress on the development
and opening of corrosion crack is small. Crack width in
PC-2 was approximately equal with PC-4.
2.2 Analytical Investigation
2.2.1 Modeling of Corrosive Expansion of PC
Tendon
The three dimensional finite element program Diana
[7] is employed to simulate corrosion crack in concrete
member, in which a non-linear strain-softening
behavior in crack propagation by Reinhardt [8] is
considered. It is assumed that the PC tendon expands
due to corrosion as shown in Fig. 5 and Eq. (2):
n(
0
+d)
2
-n.
0
2
y-1
= n. r
0
2
- n(r
0
-r
Ioss
)
2
(2)
where, dr is increment of radius of PC tendon due to
corrosive expansion under unrestraint condition (mm),
r
Ioss
is radius loss of PC tendon (mm), r
0
is original
radius of PC tendon (mm) and is expansion
coefficient of rust which the ratio between volumes of
corrosion product to the volume of PC tendon
consumed in the corrosion process. The value of was
determined as 3.14 based on the experiment by Zhao
[9]. PC tendon was modeled as an elastic material.
Radius loss of PC tendon was converted into mass
loss [10] by following equation:
w
Ioss
=
|n.
0
2
-n(
0
-
lcss
)
2
].p
s
2n.
0
=
|2
0.
lcss
-
lcss
2
].p
s
2
0
(3)
where, W
loss
is amount of corroded PC tendon in
gr/mm
2
and
s
is density of PC tendon (7.85 10
-3

gr/mm
3
).
2.2.2 Finite Element Analysis
The finite element code Diana-version 9.4.4 was
used in numerical analysis of specimen. The
stress-strain relationship shown in Fig. 6 is used for
concrete in tension to express crack propagation [8].
The constant value c
1
is 0.31. The softening branch is
defined by the following equation:
t
=
L
1-_
s
cr
s
ult
]
c
1
(4)
Uniaxial tensile strength (f
t
), Youngs modulus (E)
and fracture energy (G
f
) were estimated from
compressive strength and maximum aggregate size as
stated in CEB model code [11]. The material properties
applied are presented in Table 3.
The specimen was discretized using eight-node
isoparametric solid brick element as shown in Fig. 7.
External compressive load was applied in longitudinal
direction to simulate pre-stressing PC

Fig. 5 Corrosive expansion in cross section of PC tendon.

Fig. 6 Constitutive model of concrete in tension.

Table 3 Properties of materials used in the analysis.
Properties
Material
Concrete PC tendon
Young modulus, E (MPa) 34,070 201,000
Poissons ratio, 0.2 0.3
Tensile strength, f
t
(MPa) 3.01 2,018
Fracture energy, G
f
(N/mm) 0.1417 -

Fig. 7 Meshing for specimen to investigate of the effect of
compressive stress.
r
0
r
loss
dr
G
f
/h
f
t
cr
ult
t
(a) Front view (a) Side view

1526
tendon. Red
considered.
Analytica
averaging cr
experimenta
relative disp
crack. Expa
mass loss Eq
2.2.3 Ana
Crack pat
appeared is
on the cente
Crack wi
corrosion of
specimen, an
to corrosion
surface of co
the experim
stress contri
pre-stress on
there is not m
PC-2 and PC
Comparin
at the same

Fig. 8 First

Fig. 9 Nume
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0
C
r
a
c
k

w
i
d
t
h

(
m
m
)
Simulat
duction of pre
al crack wid
rack width a
al work. Crac
placement bet
ansion of PC
qs. (2) and (3
alytical Result
ttern in specim
shown in Fig
r of the speci
idths of the
f PC tendon
nalytical crac
n of PC tendo
oncrete. It is
mental results
ibutes to the
n corrosion
much differen
C-4.
ng analytical
e level of co
crack on surfa
erical result of
2 4
Corroded P
Analytical, PC-0
Analytical, PC-2
Analytical, PC-4
tion of Corros
estress due to
dth was als
at three same
ck width was
ween the two
tendons was
).
ts and Discus
men PC-2 wh
g. 8. Corrosi
imen.
specimens a
are showed
ck width is alm
on after it wa
found in Fig
s in Fig. 4, t
crack width
crack is ver
nce among th
and experime
orrosion, exp
ace of specimen
crack width.
6 8
PC tendon ( x 10
0
2
4
sion Crack in
o corrosion is
so evaluated
e points with
s calculated f
o nodes across
s calculated f
ssions
hen the first c
on crack app
as a function
in Fig. 9. In
most proporti
as initiated on
g. 9, as well a
that compres
but the effec
ry small bec
e results of P
ental crack w
perimental cr
n PC-2.
8 10
0
-4
g/mm
2
)
n Pretensione
s not
d by
h the
from
s the
from
rack
pears
n of
any
ional
n the
as in
ssive
ct of
ause
C-0,
width
rack

wid
gen
asso
exp
gel
of c
redu
tend
app
inte
Fig
first
thou
crac
resu
can
pro
sim
corr
3. I
of P
3.1
T
corr
and

Fig.
of P

Fig.
arra
12
ed Prestresse
dth is smaller
neration of
ociated with
pansion of PC
which can pe
concrete. Ther
uced and ma
don. It is ob
peared later
eresting obser
s. 4 and 9 is
t crack appe
ugh there is
ck width be
ults, the slope
n be consider
duct on cra
mulated by lin
rosion crack r
Investigatio
PC Tendon
Analytical In
The influence
rosion cracki
d the finite ele
. 10 Specimen
PC tendon.
. 11 Mesh
angement of PC
C1
C2
C3
S1
S2
a.Crosssecti
(a) Cross sec
(a) Side view
ed Concrete M
r than analyt
corrosive p
h corrosion
C tendon. In re
enetrate into
refore, pressu
ay not be pro
bserved that
than the an
rvation which
the increasin
eared on su
much differe
etween anal
e of curves is
red that the
ack opening
near expansio
reaches the c
on of the E
n on Corros
nvestigation
e of the arran
ing is invest
ement meshin
n to investigat
hing for to
C tendon.
C2
ion
100
tion
Members
tical one. Thi
product of
was simulat
eality, corrosi
concrete pore
ure by corrosi
oportional to
experimenta
nalytical resu
h can be con
ng of crack w
urface of con
ence in absol
ytical and
s close with e
contribution
g can be ap
on of PC tend
oncrete surfa
ffect of Ar
sion Crack
ngement of PC
tigated based
ng is shown in
te the effect of
investigate t
b. Front V
500
PC
300
(b) Front v
(b) Front
is is because
PC tendon
ted as solid
ive product is
es and cracks
ive product is
corroded PC
al first crack
ult. Another
nfirmed from
width after the
ncrete. Even
lute value of
experimental
each other. It
of corrosive
pproximately
don after the
ace.
rangement
k
C tendons on
d on Fig. 10
n Fig. 11.
f arrangement
the effect of
View
Tendon 46
100
view
view
e
n
d
s
s
s
C
k
r
m
e
n
f
l
t
e
y
e
t
n
0
t
f

The thick
cover (C3) a
that crack a
occur. S1
perpendicula
Horizontal
between PC
specimens
element prog
with paragra
3.2 Results a
Before cr
corrosive ex
cracks in co
Fig. 12.
In specim
and four PC
inner cracks
crack can be
is due to th
specimen II
inner cracks
horizontal s
Neither hori
in specimen
tendons.
Fig. 13 sh
surface of s
surface in sp
increasing o
more. This l
specimen II
appeared on
only one cra
and III. Sinc
parallel to t
short, the c
tendons can
magnified b
line.

Simulat
knesses of s
are set greate
at the top sur
signifies
ar to surfa
space, which
C tendons i
listed in Ta
gram and con
aph 2.2.
and Discussio
rack appears
xpansion is
oncrete aroun
men I, concrete
C tendons are
. Connection
e observed in
he short dista
I, horizontal
s. On the con
spaces are
zontal nor ve
n IV due to
hows appearan
specimens. A
pecimen I and
of corrosion
localization o
I and IV. As
n the surface
ack opening
ce the distance
the surface, i
corrosive exp
form one un
y concentrate
tion of Corros
ide covers (
er than the to
rface of spec
vertical sp
ace, between
h is parallel
is represente
able 4 are
nstitutive mo
ons
on the surf
consumed to
nd the PC ten
e between PC
connected w
n between PC
n specimen II
ances between
spaces are n
trary, in case
connected b
rtical inner cr
o the large
nce of corros
After crack
d III, some cra
and others d
of cracks was
s results, two
in specimen
on the surfa
e of horizonta
in specimen
pansion of t
nified crack.
ed corrosive
sion Crack in
(C2) and bot
op cover (C1
cimen is likel
ace, which
n PC tend
l to the surf
ed by S2. F
analyzed. F
dels are comm
face of conc
o generate in
ndon as show
C tendons crac
with each othe
tendons by in
and III too.
n PC tendon
not connected
e of specimen
by inner cra
racks did dev
distance of
sion cracks on
appeared on
acks opened w
did not open
s not observe
o opening cr
II and IV, w
ace in specim
al space, whic
I and III is v
two adjacent
Crack width
expansion in
n Pretensione
ttom
1) so
ly to
is
dons.
face,
Four
inite
mon
rete,
nner
wn in
cked
er by
nner
This
s. In
d by
n III,
acks.
velop
the
n the
the
with
any
ed in
acks
while
men I
ch is
very
PC
was
one
Tab
arra
Spe
I
II
III
IV

Fig.
surf

Fig.
ed Prestresse
ble 4 Dimensi
angement of PC
ecimen C1 (m
15
15
15
15
. 12 Inner cr
face.
. 13 Crack pa
surface
cracks
a. Specime
cracks
surface
c. Specime
(a) Specimen
(c) Specime
(a) Specimen
(c) Specimen
ed Concrete M
on of specimen
C tendon.
mm) C2 (mm) C
80 8
80 8
80 8
80 8
ack patterns j
atterns on the s
e
en I
e
en III

n
n

Members
ns to investiga
C3 (mm) S1 (m
80 20
80 20
80 80
80 80
ust before cra
surface of spec
surfac
cra
b. Spec
surf
cra
d. Speci
(b) Specime
(d) Speci
(b) Sp
(d) Sp
1527
te the effect of
mm) S2 (mm)
20
80
20
80
acks appear on
cimens.
ce
acks
cimen II
face
acks
imen IV
en
imen
pecimen
pecimen
7
f
n

1528

Fig. 14 Width of the largest opening crack in each
specimen as function of corrosion of PC tendon.

Fig. 14 shows the width of largest opening crack in
each specimen as a function of corrosion of PC tendon.
As a result of localization of opening crack, the greatest
crack width is developed in specimen I. Crack width is
then followed by specimen III, IV and II.
4. Conclusions
Based on the results of the experimental and
analytical investigation on corrosion crack in
pretensioned PC members, following conclusions are
obtained.
It was confirmed by both experiment and analysis
that compressive stress in the longitudinal direction
accelerates development and opening of corrosion
crack. However, the degree of acceleration was very
small. It is, therefore, that compressive stress in
concrete by prestressing does not affect significantly
on the development and opening of corrosion
crack.
When PC tendons in pretensioned PC members are
arranged with small spatial intervals, wider corrosion
crack may be developed on the concrete surface. This
is because PC tendons are connected with each other
by inner cracks and consequently opening crack is
localized. In particular, short vertical interval between
PC tendons causes wide surface crack.

Acknowledgments
The authors acknowledge the support of the concrete
laboratory at Nagaoka University of Technology,
Japan. A financial support to continue this research
from Indonesian government is also greatly
acknowledged.
References
[1] N. Nakamura, H. Mutsuyoshi, H. Minh, H. Taniguchi,
Effect of filling grout in the corrosion degradation of PC
beam, Proceeding of the Japan Concrete Institute 27 (2005)
661-666.
[2] T. Miyagawa, Y. Suzuki, T. Kondo, T. Yamamoto,
Experiment of corrosion crack on the surface of PC
member, Proceeding of the Japan Concrete Institute 33 (1)
(2011) 1091-1096.
[3] Y. Tanaka, T. Yamaguchi, S. Kuga, K. Ibayashi, An
experimental study on shear bending performance of the
pretensioned PC attacked by corrosion in PC steel,
Proceeding of the Japan Concrete Institute 33 (1) (2011)
1127-1132.
[4] F. Li, Y. Yuan, C.Q. Li, Corrosion propagation of
prestressing steel strands in concrete subject to chloride attack,
Construction and Building Material 25 (2011) 3878-3885.
[5] K.K. Tran, H. Nakamura, M. Kunieda, N. Ueda, Three
dimensional behaviour of concrete cracking due to rebar
corrosion, Procedia Engineering 14 (2011) 419-426.
[6] H. Minh, H. Mutsuyoshi, N. Nakamura, N.D. Hai,
Influence of grouting conditions on deterioration of post
tensioned prestressed concrete beams, Japan Concrete
Institute 28 (1) (2006) 773-778.
[7] DianaFinite Element Analysis, Users Manual, Release
9.4.4, TNO DIANA, BV, Delft, The Netherlands, 2010.
[8] H.W. Reinhardt, Fracture mechanics of an elastic softening
material like concrete, Heron 29 (2) (1984) 1-35.
[9] Y. Zhao, H. Ren, H. Dai, W. Jin, Composition and
expansion coefficient of rust based on X-ray diffraction
and thermal analysis, Corrosion Science 53 (2011)
1646-1658.
[10] Y. Saitou, T. Shimomura, Effect of non-uniformity of
corrosion of steel in concrete on criteria of corrosion crack,
Proceeding of the Japan Concrete Institute 30 (1) (2008)
1089-1094.
[11] CEB-FIP Model Code, Thomas Telford Ltd., Comite
Euro-International du BetonFederation Internationale de
la Precontrinte, London, 1990.

0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 2 4 6 8 10 12
C
r
a
c
k

w
i
d
t
h

(
m
m
)
Corroded PC tendon (x 10
-4
g/mm
2
)
SpecimenI
SpecimenII
SpecimenIII
SpecimenIV
Framework for Quality Improvement of Infrastructure
Projects
Abukar Warsame
Department of Real Estate and Construction Management, School of Architecture and Built Environment, Royal Institute of
Technology (KTH), Stockholm 100 44, Sweden

Abstract: In order to achieve high quality that not only gives acceptable return value to society but also satisfies the needs of all the
stakeholders of infrastructure projects, comprehensive understanding of issues pertaining to the quality of the project is needed. The
aim of this study is to provide an overview the most common procurement methods used in constructing infrastructure transport
projects and analyze how these methods contribute to the desired quality of the final product in relation to client competence. An
on-line survey of construction actors was carried out to ascertain quality level of Swedish infrastructure transport projects and
determinant factors of quality problems. An equal number of respondents indicated that the quality of infrastructure projects has
either increased or remained same level over the past twenty years. Respondents also pointed out lack of client competence that is
vital in realizing the desired quality level through proper procurement, monitoring and evaluation procedures. Public clients heavily
rely on traditional design-build procurement that requires considerable client involvement of a project. Thus, the association of
quality problems and lack of client competence may not be a mere coincidence but an overlooked outcome of current situation.

Key words: Competence, infrastructure transport, procurement, quality, warranties.

1. Introduction

Transport infrastructure transport projects are
essential to economic activity and growth of a society.
They are characterized as having long-life and long
gestation processes as well as being capital intensive
[1, 2]. The sheer size of transport infrastructure
projects often necessitates government involvement in
terms of financing, developing, operating and
maintaining them. Several arguments that explain
government involvement in delivering infrastructure
transport were outlined by Ref. [3]. One of the
arguments is divergence of interest between public
and private sector that makes transport infrastructure a
product that cannot be easily delivered by the private
sector.
However, government involvement in procuring
infrastructure transport projects has been reduced for a

Corresponding author: Abukar Warsame, Ph.D., research
fields: real estate economics and construction management.
E-mail: abukar.warsame@abe.kth.se.
lot of reasons while private sector participation has
been on the rise lately [1, 4]. Private sector
involvement in developing major infrastructure
transport projects such as roads and railways were
often motivated by the financial constraints faced by
the public sector [4]. The private sector may provide
more efficiently and faster delivery of infrastructure
transport than public sector because of higher
management expertise and autonomy as well as lower
levels of regulation and control that is associated with
private sector [2]. Private sector efficiency allows
government to share the associated risk of asset
procurement and service delivery with the private
sector [5]. Development of new products and
technologies that lead to cost reduction and quality
improvement also contributed to increased private
sector participation [5] even though cost-cutting and
desire of better quality are sometimes difficult to
achieve both of them simultaneously [6].
Some of the causes of project success and failure
can be traced back to the procurement method and
DAVID PUBLISHING
D
Framework for Quality Improvement of Infrastructure Projects

1530
how the owners selected that method [7, 8]. The
owners or clients are the most critical to project
success [9] and their characteristics such as expertise
and experience are the moderating factors on the
performance procurement system chosen [10]. A UK
government report [11] claims that performance in
the successful delivery of outcomes is strongly
dependent on the skills of the client, not simply on the
contract structure.
The Swedish construction industry is facing a
shortage of skilled workers and ageing of existing
staff [12]. Since quality performance and the
effectiveness of any selected procurement method
partially hinges on the competence of clients
workforce, it is crucial to examine how different
procurement methods could be affected by the scarcity
of client competence, experience and expertise. The
aim of this paper is to examine the most dominant
procurement methods used in constructing
infrastructure transport projects and the role that client
competence played on achieving the desired quality of
the final product. Identifying some of factors that are
associated with quality problems, i.e., lack of
competence and at what stage of the construction
process that these factors are critical will allow
procurement officers and decision makers to
contemplate which procurement method is appropriate
to certain situations.
2. Literature Review
A survey conducted by the International Federation
of Consulting Engineers [13] found that decreasing
quality of construction is a worldwide phenomenon
that is mainly caused by inappropriate mechanism of
project delivery such as poor consultant and contractor
selection, bad design, poor project supervision and
inadequate material and workmanship. Other authors
[3, 14, 15] identified several factors that have
significant influence on the quality performance of a
project however one common denominator of all these
inadequacies or quality problems is the lack of client
competence. Project managers competence, top
management support and interaction between projects
participants are external contributors that enhance the
project quality performance from its existing level
while owners competence is an internal contributor in
the sense that owners tend to retain the quality
performance at the existing level itself [3, 14].
A competent project manager and project team
were highly ranked among the top five project success
factors in a study carried out by Nguyen et al. [15].
Client incompetence often were blamed on flawed
expectations and requirements placed on contractors
[16]. Respondents in Ref. [16] survey pointed out
incompetent client were unaware of the requirement
they should or did in fact place on the supplier and
thus contractors were inclined towards not to
following these requirements. The FIDIC [13] finding
also highlights the importance of client competence
during the planning and design stage as well as during
the tendering and construction stages. It emphasizes
how the actors from the demand-side of the
construction process such as client and engineer
(consultant) could play a big role in improving the
quality of construction.
During the operation and maintenance phase,
assessment of the quality is challenging since the
expected life of many infrastructure projects is quite
long. Customer satisfaction surveys are sometimes
used as a proxy for the quality level of a product.
Though this approach can give an indication of quality
level at certain point in time, it could be difficult to
objectively assess the quality level of road
infrastructure from end-users point of view. A survey
of Swedish Transportation Administration [17]
exemplifies the difficult of ascertaining reliable
measurement of satisfaction level from road users.
The survey found that drivers were not only
dissatisfied with road conditions but also private and
professional drivers have different opinion about the
level of dissatisfaction. Many road administrations
from the countries that participated in this study

1531
reported shrinking staff due to retirement and to the
attractiveness of private sector employment. This
report also has noted that new types of contracts
require broader project management skills and
practical experience and thus there is need for key
competences such as operations, road surfacing and
specialist engineers [17]. Further discussion on the
implication of client competence in relation to the
quality of infrastructure projects will be covered later
(Section 5). However, the most common procurement
methods and how they defer in terms of quality aspect
will be discussed next.
3. Procurement Methods
Procurement methods are often classified based on
how construction activities such as design,
construction, and operation and management are
delegated among actors in the project. Different
financing options of infrastructure projects also
influence this classification. According to Ref. [18],
procurement systems are classified as: traditional
(separated), design and construct (integrated),
management (packaged) and collaborative (relational).
Short description of procurement methods, which is
similar to Ref. [18] classification, will be described as
follows.
3.1 Traditional Procurement (Separated)
Traditional procurement method of DBB
(design-bid-build) with negotiated or separated
competitive bidding fits this type of classification.
DBB procurement method is prescriptive by nature
where owners with in-house designers or with
appointed consultant prepare the project design and
tender documents. QC (quality control) and QA
(quality assurance) activities are either carried out
altogether by the contractor or sometimes the client is
responsible for quality assurance while contractor is
responsible for quality control process. In Sweden,
twenty-five years ago, the client (Swedish
Transportation Administration) has transferred quality
control and quality assurance to the contractors and
accepts only the final project if it meets the expected
quality level of performance.
Mandell and Nilsson [19] studied some 1,400 road
construction and renewal projects procured by the
Swedish road administration between 2000 and 2009.
The large majority of these projects were procured in
the form of unit price contracts or DBB. Pietroforte
and Miller [20] noted that DBB was the most
dominant form of infrastructure procurement after the
Second World War in the US although during the last
decade DB (design and build) has grown steadily in
both private and public sector. Unit price contracts
dominate procurement of projects in the transport
sector in Sweden [19] and in US [21]. One
shortcoming of DBB approach is that it does not take
into account the increasing operation and maintenance
costs of the ageing infrastructure [20]. Furthermore,
DBB procurement focuses the price of the project
rather than quality.
3.2 Management (Packaged)
CM (construction management), management
contracting, and design and mange belong to this
category. Basically, an owner enters agreement with a
construction firm that carries out leadership,
administration and management of specified services
[22]. According to Ref. [22], CM is seldom used in
road construction because of the relatively small
number of contractors involve in road construction
that is easily manageable by clients.
3.3 Design and Build Procurement (Integrated)
In order to address certain problems associated with
DBB and improve the performance of construction
projects, integrated DB (design-build) procurement
and its variants such as DBO/M
(design-build-operate/maintenance) have been
promoted to be an appropriate alternative procurement
strategy. DB is an integrated procurement system
where the client contracts with a single contracting

1532
organization to carry out both design and construction
responsibilities with or without inclusion of operation
and maintenance contracts. One common feature of
these types of contracts is that client is responsible for
entirely and directly financing the project [20]. DB
would allow contractors to tender the most
economical design that meets the requirements of the
client and use materials and innovation techniques
that produces desirable outcome for the client.
However, DB procurement method demands different
skills and competences than the traditional
procurement (DBB) that public servants are
accustomed [22]. A distinct criterion for DB contracts
is the RFPs (requests for proposals) along the price
and technical proposals. The RFPs contains owners
objectives and needs with respect to quality and the
design/builders are required to interpret those
requirements and submit their proposals [23].
There is no definite agreement on whether the
overall quality achieved under DB procurement is
better than DBB. Clients expectation of design
quality does not differ between DB and DBB though
contractors under DB tend to choose only the
necessary design that leads to savings [24]. One major
problem found by Ref. [23] is that owners of public
projects who have used DB procurement often
executed with DBB mentality because majority of
RFPs contained a DBB construction quality control
plan requirement. As Ref. [25] argue, the use of DB
procurement does not mean that an inexperienced
client can simply leave all the project and
responsibility to the DB contractor. Design-build
procurement is much easier for a client with sufficient
design and past construction experience [25].
3.4 Public-Private-Partnerships and Its Variants
(Relational)
A further categorization of procurement methods is
based on how financing of the project is formulated.
This category is an extension of integrated
design-build variations except that the private sector is
now financially involved in the provision of a project.
Design-build-finance-operate or BOT, as it is known
outside the US [20], BOOT
(build-own-operate-transfer), and BOO
(build-own-operate) belong to this group. These
procurement methods have longer contract period than
the typical integrated procurement methods financed
by the publicly sector. However, there is no clear-cut
definition between PPPs (public-private-partnerships)
and BOOT except that one can observe an increase in
the number of stakeholders of PPP projects [26].
DBFO (design-build-finance-operate) or BOT
(build-operate-transfer) is another example of PPPs
arrangement.
Though the use of PPPs for procurement of public
infrastructure projects is not something new to most of
the developed and developing countries,
comprehensive performance assessment of the
projects undertaken with PPPs has been difficult due
to the divergence between the duration of PPPs
contracts and the longevity of infrastructure projects.
Infrastructure projects could last decades or centuries
while PPPs contracts are often in the range of 30 years.
Thus, it is difficult to have a full assessment of
infrastructure projects that are constructed under PPPs
arrangements. Procurement of infrastructure projects
with PPP involves many parties with conflicting
objectives and thus requires extensive client
competence and expertise [5]. Private operator may
heavily invest in cost reduction technologies without
taking into consideration their impact on the quality of
the project [6]. Realized quality could be better or
worse under private contracting depending on the
impact of any cost reduction technologies employed
by the private actors [6]. PPP arrangements to surface
transport infrastructure are complex with many pitfalls
and thus require strong client competence and
expertise [27]. Public sector experience with the
design and build procurement provides adequate
knowledge and capacity that is needed to handle
complex PPPs arrangements [27]. Based on

1533
international experiences, Swedish Transport
Administration report [28] accentuates the importance
of highly competent client organization for efficient
and successful PPP projects. According to Ref. [28],
the only infrastructure project that has been built with
PPPs arrangements so far is Arlanda railway link
(Arlandabanan).
4. Research Method
After a pilot survey with a number of practitioners
and academics in the field of construction, an on-line
questionnaire containing 29 questions was sent to 128
selected respondents (45 contractors, 35 clients, 37
consultants and 11 regional traffic offices). The
composition of respondents was intended to ensure a
fair representation of different stakeholders of
construction projects that would allow us to ascertain
the extent of quality concerns in the sector. Through
the authors research reference group and organization
websites as well as professional associations websites,
the appropriate persons in each organization that could
answer the questionnaire were identified. After few
questions related to the information of respondents,
respondents were asked their general assessment of
the quality problems today and the past. In order to
establish the prevalence of quality problems in
relation to different phases (construction phase, final
inspection phase and during warrant period),
respondents were asked their opinion about how often
quality deficiency occurs and who discovers as well as
what actions are taken when quality problems are
discovered.
Warrant contracting is usually used with
performance-based specifications and is intended to
replace stringent quality control and inspection
regimes associated with traditional procurement [29].
When quality deficiency is discovered three possible
actions can be taken by project managers: complete
removal of the defective product or structure, repair
without complete removal, and deductions from
payments if the quality is not exactly the required
level but acceptable within certain limits. A set of
questions that was intended to give people a sense of
understanding on whether warranty contracting has
improved quality of construction projects after the
client transferred quality assurance responsibility to
the contractor were also included in the questionnaire.
The rest of the questions of the survey (15
questions) were statements that are intended to
ascertain important attributes such as competence of
construction actors, project characteristics etc. that
may have contributed quality problems or lack of
quality improvement. Respondents were also asked to
comment on each question and statement in order to
solicit their candid view about the quality of
infrastructure projects.
5. Frame Work for Quality Improvement
5.1 Construction Quality
Brockmann [30] argues that construction projects
such as buildings or infrastructure are typically
quasi-credence goods meaning that their qualities are
ex-ante intangible but ex-post tangible. His argument
is based on the classification of three types of quality
[31], search qualities which are known before
purchase, experience qualities which are known
costlessly only after purchase, and credence qualities
which are expensive to assess even after purchase.
Construction goods do not exhibit search qualities
since the quality of construction goods can not be
determined at the time of signing contract [30].
However, once the contract is carried out and the
project is completed ex-post search qualities could
become tangible. Similarly, construction goods also
differ from experience goods since experience
qualities are based on a high frequency of contracting
between the same client and contractor [30]. The
nature of construction business and government
regulations as well as competition rules does not
foster repetitive interaction among actors in
construction sector. On the contrary, Vassallo [32]
argues that quality of most of infrastructure projects

1534
are observable after their use and classified them as
experience goods. He claims that infrastructure
quality is verifiable though the cost of measuring
quality is not usually low.
It could be argued that since different procurement
methods provide construction clients an opportunity to
influence design and construction qualities, a clients
internal capacity and available resources such as
design and inspection teams as well as his competence
and skills play a major role determining whether the
desired product has the characteristics of search,
experience or credence qualities. If a client or his
representative has the capability to carry out all the
necessary quality control and quality assurance
activities, this will ensure that the quality of the
project becomes not only observable but verifiable
during the construction until the operation phase starts
or even during the warrant period. When client and
contractor have also long term relationship high
frequency contracting, which is a necessary condition
for experience qualities [30], could be materialized.
The quality and long term performance of the
infrastructure is subject to many external and
stochastic factors such as the level of usage of
infrastructure, efficiency of intended operation, weather,
and frequency of planned and unplanned maintenance.
Thus, the quality of realization during operation and
maintenance is may be what makes construction
goods to be treated as quasi-credence goods.
5.2 Client Competence
A client organization with highly skilled and
sufficiently experienced workforce is most likely
capable to ascertain many project risk factors than
client organization with less endowed human capital
resources. There are distinctive competencies that are
highly required when client is employing a specific
procurement method. The traditional procurement
(design-bid-build) where client has multiple contracts
with different actors for the provision of infrastructure
projects is a good starting point to identify most
relevant responsibilities of client. Client is expected to
have enough manpower resources with appropriate
skills and expertise in different areas such as financial,
technical and contracting management. Some of the
notable client competences are: ability to define
project scope and objectives, establish design criteria
and performance requirement, carry out preliminary
survey and geotechnical investigations, ensure
constructability of design, perform control and
inspection of quality performance, and prepare
possible mitigation actions and procedures when
performance objectives are not met.
Competences that are needed for risk sharing
activities where the client uses a procurement strategy
such as design-build and performance-based contracts
with warranties are much similar to the above list
except that client is oblige to prepare RFPs for DB
contracts instead of complete design of the project.
Previous experience with DB related projects were
found to be among the key competences that DB
clients in China should possess [25]. Competences
that are needed for risk transferring activities where
client engages procurement strategies such as DBFO
with or without maintenance and other forms of PPPs
are now more related to management, financial, legal
and commercial activities.
A resourceful client organization could use these
extraordinary technical competencies and expertise to
transform experience qualities to search qualities or
even credence qualities to experience qualities. This
transition between types of qualities would have
certain implication on how client selects appropriate
procurement methods. Thus, the above representation
could improve understanding the impact of different
procurement methods on clients workforces and their
skill development.
6. Results and Discussion
6.1 Quality Concerns
Sixty seven respondents that translate to a 52%
response rate completed the survey. According to the

1535
respondents, quality of construction projects
constructed in last five years is either at the same level
or even better than the quality level of projects built
twenty years ago. Same numbers of respondent (43%)
believe that quality level is higher today or has not
changed during this period.
Responses from client side were almost
inconclusive when asked who finds out quality
problems during construction (client or contractor). A
possible explanation for this inconclusiveness could
be that the client has little opportunity to discover
quality problems during this phase since quality
assurance responsibilities were transferred to the
contractor. Respondents indicate that clients
complain is limited to few isolated projects (47%)
though 21% of them pointed out that client complain
about quality problems is present in majority of the
projects. With regard to quality problems discovered
during the final inspection, 44% of the respondents
say that few isolated projects encountered quality
problems while 26% of them indicated that majority
of the projects experienced quality problems.
A combined 43 respondents (66%) indicated that
quality problems discovered during the final
inspection get fixed by means of reparations without
complete removal or contractors accepted payment
deductions. The situation is somewhat different when
it comes to quality problems during warranty period
or shortly after it expires. Majority of respondents (42)
indicate that clients very seldom discover or complain
about quality during this period. If quality deficiencies
are discovered during the warrant period, a complete
removal or payment deductions are not the preferred
options. Twenty-five respondents or 42% of them say
that very seldom contractors remove the defected part
or structure. Another 20% (20 respondents) say that
payment deductions do happen very seldom. The
prevailing measure to rectify quality problem during
warranty period is reparation without complete
removal. These responses give an indication on how
project managers and client representatives react when
they discover quality problems but the underlying fact
is that these kinds of decisions require higher
competence and experience that is in short supply in
the client organizations workforce as the following
responses suggest.
6.2 Client Competence Concerns
Responses from the questions related to
competence of client with regard to quality problems
of infrastructure projects were very strongly negative.
Respondents indicated that clients lack of
competence is major factor of quality problems of
infrastructure projects. Approximately 80% of
respondents partially or totally agree with the
statement that the quality problem in the finished
structure is highly dependent on the clients
competence. Similarly, 72% of them indicated that
tendering documents contributed quality problems of
final product. Respondents also pointed out (80%) that
designers of the project play a major role on the
quality problems experienced in the finished project.
The above finding provides an opportunity to
investigate the role of client competence on different
procurement methods and how much skills, expertise
and competence of client workforces and client
experience is demanded by each type of procurement
method.
A prognosis from Ref. [12] points out that one third
of the construction industrys workforce will retire
between 2005 and 2015. The study also shows that the
construction industry has a higher proportion of older
people and a smaller proportion of younger employees
compared to the total employment. In Finland,
employees retirement and reduction of client staff are
reported to be behind diminishing client experience
and competence [22]. The implication of human
capital scarcity is that public clients face an uphill
battle in attracting new talented and competent
graduates as well as retaining them in a competitive
market. Some of the respondents in the survey also
raised this issue of ageing client workforce,

1536
non-replacement policy of retired staff and their
concern of new skilled workers not joining the public
transport sector.
6.3 Procurement Method Concerns: Client
Competence and Quality
According to Ref. [33], no one procurement method
is likely to be better than others for any project
although one procurement method could be more
appropriate or suitable than others for an individual
project. In Sweden, traditional procurement
(design-bid-build) or unit price contract has been the
most dominant contract form used for the delivery of
transport infrastructure projects [19, 28]. The question
is whether this heavy reliance of DBB method in the
Swedish infrastructure transport is supported by
undisputed higher performance achievement
compared to other procurement methods or the
Swedish Road and Railway administration has chosen
it for other reasons.
Several possible explanations have been offered as
to why a clients or owners keep using repeatedly the
same procurement method especially the dominance
of the traditional procurement method. Familiarity of
DBB within the industry and its ability to satisfy
public accountability, client control over the projects
outcome and cost certainty makes easy and attractive
method for public sector to rely on more often than
other procurement methods [24, 33]. Avoidance of
uncertainty is another explanation. Ldre et al [8]
claim that owners select a well-known procurement
route since unknown procurement method could
introduce new uncertainty. When owners attain
experiences from the use of certain procurement
procedure and management routines, it will encourage
them to keep using this combination in their next
project [8]. DBB procurement are generally
considered to be suitable when client wants to settle
upon a design before construction commitments, take
advantage of existing designs or the client has the
only experience necessary for this kind of project [22].
Public owners are not motivated to be creative when
selecting the procurement route because creative
thinking is seldom rewarded while project failure will
be criticized [8]. Thus, it is convenient to use the same
procurement system. One important question that
could be raised about the heave reliance of public
sector on DBB procurement method is can the
benefits and confidence that a continuous use of DBB
offers to the public sector be sustainable for a long
period of time?
While cost effectiveness, strong client control of the
project and flexibility are some of the benefits
associated with DBB, one can not overlook the
downside of the use of DBB procurement for several
reasons: First, if the clients lack of competence (as
the survey indicates) is due to a shortage of skilled
and experience workforce then the use of DBB could
exacerbate the situation since DBB requires higher
client involvement; Secondly, the use of DBB
contracts has been seen as one of the factors that
contributed to the lower productivity and lagging
performance of the Swedish construction industry
[19, 34]; thus, a lack of client competence and its
negative impact on quality of infrastructure projects
will worsen the situation and inflict further distress to
the sagging industrys productivity.
The frequent use of DBB procurement raises other
concerns on top of its susceptibility for human capital
shortages. In US, economic factors and change of
procurement laws will cause an increase of public
clients use of other procurement methods such as DB,
DBO and BOT [20]. Similarly, since April 2000 UK
government requires projects should be procured by
PFI (public-financing-initiative), prime contracting or
design-build [8]. In Sweden, limited projects have so
far been procured with other procurement method than
traditional DBB method. However, the use of
performance-based contracts such as DB (with
short-term warranties) and DBOM (with long-term
warranties) is painted as good move toward public
sector readiness to embrace PPPs arrangements [28].

1537
The use of warranty contracting has given some
leeway for the public sector to deal with quality
problems that could arise from contractors not
complying with the specifications and the design
[29, 35]. Warranties has offered an alternative way to
assure performance when State highway agencies
faced staff and budget shortage and still needed to
increase the quality and life-cycle performance of
pavement [35]. Warranties guarantee that contractors
are responsible to repair and replace defects both
during the construction and warranty period [29]. One
of the main features of warranties is that quality is
measured based on actual product performance over
time rather than construction materials and
workmanship [36]. In other words, contractor has the
incentive to use any construction methods and
products as long as they meet clients specified quality
performance. The finding of the survey seems to
support that contractors with warranty obligations
have actually succeeded to produce the desired quality
performance of infrastructure projects. More than
seventy percent of respondents have indicated that
only few or no infrastructure projects encountered
quality problems during the warranty period or after
the warranty expired (after 1 to 3 years).
7. Conclusions
The success and failure of infrastructure project to
achieve their performance objectives in terms of
end-users needs and societies economic benefits are
determined by number of factors including the
procurement strategy. The choice of an appropriate
method to procure a specific infrastructure project
depends on many factors and client competence is one
of the most important of them. The client is not only
the owner of the project and the initiator of the
concept but also represents the end-user and thus
responsible for determining their needs objectively,
interpreting them accurately and selecting design and
construction teams that can deliver successfully the
desired product. All these activities and
responsibilities require a very strong client
competence with skills, expertise and experiences
necessary to carry out their technical, financial and
management duties.
As the findings of the survey and other previous
studies mentioned in this paper indicate, lack of client
competence is one of the factors that contribute
quality problems of infrastructure projects. Traditional
procurement method, which is the most common
method used by the STA (Swedish Transportation
Administration), demands the highest client
involvement in the project compare to other
procurement methods. In light of shortage of skilled
and experienced workforce in the public sector and
the heavy reliance of the sector on this traditional
procurement, it is plausible to assume that the
association of quality problems and lack of client
competence is not a mere coincidence but an
overlooked outcome of current situation.
Many benefits that are associated with traditional
procurement method such as clients control of the
project, design flexibility and familiarity of the sector
with this method and public sectors achievement of
quality standards through warranties can not be
disregarded. However, the need to have enough public
sector staff with good skills and competencies is very
crucial to improve the quality of infrastructure
transport projects. Strong and broader client
competence would enable the public agency staff to
properly identify the needs of the customers that are
necessary inputs to determine the performance
requirement and the objectives of the project.
Increased client competence would ensure that many
unknown quality attributes are transformed to known
elements and known-unknowns will be shared or
transferred accordingly: On one hand, product with
quality attributes similar to manufacture goods (search
goods) could benefit a procurement method that relies
more on standardization and the use of more
prefabricated products; on the other hand, product
with experience and credence qualities would benefit

1538
more on the use of procurement methods that foster
frequent and long-term relationships such as
performance-based contracts and PPPs arrangements.
Furthermore, other procurement methods such as
relational (PPPs), integrated (DB) and
performance-based contracts (DBOM) require strong
client competence, skills and expertise that match
those possessed by the commercial and
business-oriented private sector.
References
[1] D. Grimsey, M. Lewis, Evaluating the risks of public
private partnerships for infrastructure projects,
International J ournal of Public Management 20 (2) (2002)
107-118.
[2] A. Ng, M. Loosemore, Risk allocation in the private
provision of public infrastructure, International J ournal of
Project Management 25 (4) (2006) 66-76.
[3] S. Rienstra, P. Nijkamp, Lessons from priavete financing
of transport infrastructure: Dutch infrastructure in the
19th century and European projects in the 20th century,
Revue conomique 48 (2) (1997) 231-245.
[4] G. Hodge, C. Greve, Public-private partnerships: An
international performance review, Public Administration
Review 67 (3) (2007) 545-558.
[5] E. Cheung, A. Chang, Suitability of procuring large
public works by PPP in Hong Kong, Engineering,
Construction and Architectural Management 17 (3) (2010)
292-308.
[6] A. Estache, A. Iimi, C. Ruzzier, Procurement in
Infrastructure: What Does Theory Tell Us?, World Bank
Policy Research Working Paper Series, 2009.
[7] J . Tookey, M. Murray, C. Hardcastle, D. Langford,
Construction procurement routes: Re-defining the
contours of construction procurement, Engineering,
Construction and Architectural Management 8 (1) (2001)
20-30.
[8] O. Ldre, K. Austeng, O.J. Klakegg, Procurement routes
in public building and construction projects, J ournal of
Construction Engineering and Management 132 (7) (2006)
689-696.
[9] K. Molenaar, A. Songer, Model for public sector
design-build project selection, J ournal of Construction
Engineering and Management 124 (6) (1988) 467-479.
[10] D.T. Luu, S.T. Ng, S.E. Chen, Parameters governing the
selection of procurement systemAn empirical survey,
Engineering, Construction and Architectural Management
10 (3) (2003) 209-218.
[11] H.M. Treasury, Infrastructure Procurement: Delivering
Long-Term Value, 2008, http://www.hmtreasury.gov.uk
/d/bud08_procurement_533.pdf (accessed Nov. 10,
2010).
[12] Competence in the Civil Engineering Sector, Final report,
FIA (Renewal in the Civil Engineering Industry), 2005.
(in Swedish)
[13] Improving the Quality of Construction,
FIDICInternational Federation of Consulting Engineers,
2004.
[14] K. J ha, K. Iyer, Critical factors affecting quality
performance in construction projects, Total Quality
Management 17 (9) (2006) 1155-1170.
[15] L.D. Nguyen, S. Ogunlana, D.T. Xuan, A study on
project success factors in large construction projects in
Vietnam, Engineering, Construction and Architectural
Management 11 (6) (2004) 404-413.
[16] A. Landin, ISO 9001 within the Swedish construction
sector, Construction Management and Economics 15 (5)
(2000) 509-518.
[17] STA (Swedish Transportation Administration), The Road
to ExcellenceAn International Benchmarking between
National Road Admistrations, 2010, http://
publikationswebbutik.vv.se/upload/5591/2010_075_the_r
oad_to_excellence_.pdf (accessed Oct. 11, 2010).
[18] P. Love, P. Davis, D. Edwards, D. Baccarini, Uncertainty
avoidance: Public sector clients and procurement
selection, International J ournal of Public Sector
Management 21 (7) (2008) 753-776.
[19] S. Mandell, J .E. Nilsson, A Comparison of nit price and
fixed price contracts for infrastructure construction
projects, Working Papers, Swedish National Road &
Transport Research Institute (VTI), 2010.
[20] R. Pietroforte, J . Miller, Procurement methods for US
infrastructure: Historical perspectives and recent trends,
Building Research & Information 30 (6) (2002) 425-434.
[21] D. Gransberg, C. del Puerto, D. Humphrey, Relating cost
growth from the initial estimate to design fee for
transportation projects, J ournal of Construction
Engineering and Management 133 (6) (2007) 404-408.
[22] T. Koppinnen, P. Lanhdenper, The Current and Future
Performance of Road Project Delivery Methods, VTT,
Finland, 2004, http://www.vtt.fi/inf/pdf/publications
/2004/P549.pdf (accessed Nov. 15, 2010).
[23] D. Gransberg, K. Molenaar, Analysis of Owners Design
and Construction Quality Management Approaches in
Design/Build Projects, J ournal of Management in
Engineering 20 (4) (2004) 162-169.
[24] T. Koppinnen, P. Lanhdenper, Road sector experience
on project delivery methods, VTT, Finland, 2004,
http://www.vtt.fi/inf/pdf/tiedotteet/2004/T2260.pdf (accessed
Nov. 15, 2010).
[25] B. Xia, A. Chan, Key competences of design-build clients

1539
in China, J ournal of Facilities Management 8 (2) (2010)
114-129.
[26] M. J efferies, W.D. McGeorge, Using public-private
partnerships (PPPs) to procure social infrastructure in
Australia, Engineering, Construction and Architectural
Management 16 (5) (2009) 415-437.
[27] Transport Infrastructure Investment: Options for
Effeciency, OECD (Organization for Economic
Co-operation and Development) report, International
Transport Forum, Transportation Research Center, 2008.
[28] A Swedish Model for PPP in Infrastructure Investment,
STA (Swedish Transportation Administration), 2008,
http://epppc.hu/documents/cikkek945aswedishmodelforp
ppininfrastructureinvestment.pdf (accessed Sep. 15,
2010).
[29] Q. Cui, M.E. Bayraktar, M. Hastak, I. Minkarah, Use of
warranties on highway projects: A real option perspective,
J ournal of Management in Engineering 20 (3) (2004)
118-125.
[30] C. Brockmann, Global constrution markets and
contractors, in: L. Ruddock (Ed.), Economics for the
Modern Built Environment, Taylor and Francis, London,
2009.
[31] M. Darby, E. Karni, Free competition and the optimal
amount of fruad, J ournal of Law and Economics 16 (1)
(1973) 67-88.
[32] J . Vassallo, Implementation of quality criteria in
tendering and regulating infrastructure management
contracts, J ournal of Construction Engineering and
Management 133 (8) (2007) 553-561.
[33] P. Love, R. Skitmore, G. Earl, Selecting an appropriate
procurement method for a building project, Construction
Management and Economics 16 (2) (1998) 221-233.
[34] J .E. Nilsson, The value of puplic-private-partnerships in
infrastructure, Working Papers, Swedish National Road
& Transport Research Institute (VTI), 2009.
[35] FHWA (Federal Highway Agency) Highway quality
compendium: FHWA-IF-07-012, 2007, http://
www.fhwa.dot.gov/construction/pubs/hif07012/07012.pd
f (accessed Sep. 15, 2010).
[36] k. Guo, E. Minchin, T. Ferragut, The shift to warranties
and performance specifications: What of method
specifications?, Construction Management and
Economics 23 (9) (2005) 953-963.


Reconstruction of an Experts Decision Making
Expertise in Concrete Dispatching by Machine Learning
Mojtaba Maghrebi
1
, Claude Sammut
2
and Travis Waller
1, 3

1. School of Civil and Environmental Engineering, The University of New South Wales (UNSW), Kensington NSW 2052, Australia
2. School of Computer Science Engineering, The University of New South Wales (UNSW), Kensington NSW 2052, Australia
3. National Information and Communications Technology Australia (NICTA), Sydney 20000, Australia

Abstract: Finding the optimum solution for dispatching in concrete delivery is computationally intractable because it is a NP-hard
(non-deterministic polynomial-time hard) problem. Heuristic methods are required to obtain satisfactory solutions. Inefficiencies in
mathematical modeling still make concrete dispatching difficult to solve. In reality, complex dispatching systems are mostly handled
by human experts, who are able to manage the assigned tasks well. However, the high dependency on human expertise is a considerable
challenge for RMC (ready mixed concrete) companies. In this paper, a logical reconstruction of an experts decision making is
achieved by two machine learning techniques: decision tree and rule induction. This paper focuses on the expert dispatchers
prioritization of customer orders. The proposed method has been tested on a simulation model consisting of a batch plant and three
customers per day. The scenarios generated by the simulation model were given to a dispatch manager who was asked to prioritize the
customers in each day. The scenarios and the decisions were then input to the machine learning programs, which created
generalizations of the experts decisions. Both decision trees and rules approach 80% accuracy in reproducing the human performance.

Key words: Machine learning, logic reconstruction, experts behavior.

1. Introduction

Concrete is the single most widely used material in
the world [1]. Due to technical requirements, usually
concrete must be mixed in a batch plant and hauled to
the construction site. Typically, the capacity of a RMC
(ready mixed concrete) company is limited by the size
of its fleet, hence the importance of truck dispatching
for RMCs. Optimization is the first technique that
comes to mind to solve RMC dispatching problems.
There are many examples in the literature that have
been proposed to solve this problem [2-7], however,
the RMC dispatching problem is NP-hard, which
means that there is no method that can solve this
problem in polynomial time [2, 4, 6, 7]. Thus,
obtaining optimum decisions is computationally
intractable. In the literature, approximate, heuristic
methods have been widely used by researchers to

Corresponding author: Mojtaba Maghrebi, Ph.D. candidate,
research fields: construction management and machine learning.
E-mail: maghrebi@unsw.edu.au.
achieve an acceptable result [2-4, 6-10]. In heuristic
methods, the mathematical model, which is built for
optimization is used. Even though the results achieved
by these techniques are acceptable, there are still many
problems in mathematical modelling that threaten the
reliability of the techniques. To investigate this issue
thoroughly, an RMC mathematical model that was
introduced by Ref. [2], and used by other scholars such
as Ref. [11], is described here.
RMC dispatching is represented by a network, as
Fig. 1. In this network, each customer and depot is
allocated a node. For each delivery, a connection is
needed between a depot and a customer. The amount of
concrete ordered by different customers is distinct.
The number of required deliveries is calculated for
each project, based on the ordered amounts. The size of
a triangle in Fig. 1 reflects the amount of purchased
concrete.
Although the mathematical modelling techniques
are powerful tools, there are many problems, especially

DAVID PUBLISHING
D

Fig. 1 A disp

when they a
system such
we will cov
mathematic
inefficiencie
along with th
be handled
paper is to
decisions by
will be a
automaticall
2. Insight
In this se
models in co
2.1 Large Nu
Mathemat
function tha
of the prob
constraints,
problem in
number of
constraints.
nine depots
patching netwo
are used to m
h as concrete
ver two of the
modelling. I
es of mathem
he reasoning
by human ex
o extract kn
y machine lear
able to m
ly.
into the Pr
ection, the in
oping with re
Number of Var
tical modelli
at reflects the
blem. All con
even the triv
building the
variables tha
For example
and around 2
Reconstruc
Conc
ork.
model an unce
delivery. In
e most impor
In the subseq
atical modell
behind why
xperts. The m
nowledge fr
rning and bui
make dispatc
roblem
nefficiencies
al situations a
riables
ing consists
desired goal
nditions mus
vial issues. Th
constraint m
at are needed
e, in RMC
270 deliverie
ction of an Ex
crete Dispatc
ertain dispatch
the next sect
rtant problem
quent section,
ling are discu
such systems
main aim of
om an exp
ild a platform
ching decis
of mathema
are discussed
of an objec
ls and constra
st be casted
herefore, the
model is the l
d to express
dispatching w
es with 70 tru
xperts Decis
ching by Mac
hing
tion,
ms of
, the
ussed
s can
this
erts
m that
sions
atical
d.
ctive
aints
into
first
large
s the
with
ucks,
the
Reg
prob
2.2
R
dyn
be c
data
a m
nee
such
the
han
adju
is si
Thi
syst
D
rese
lack
prac
who
them
extr
sion Making E
hine Learning
model requ
gardless of the
blem, the num
Dealing with
RMC dispatc
namic, since i
changed at an
a requires an
minor change i
eds a minor ad
h as a truck b
model. So, R
ndling requir
ustment durin
ignificant the
is take times
tem needs on
Due to the dis
earch in this f
k of practica
ctice, RMC d
o are able to
m efficiently
ract the know
Expertise in
g
ires more th
e technique th
mber of varia
h Uncertain a
ching systems
it is expected
ny time [2, 6
adjustment in
in the system,
djustment in m
reakdown, ne
RMC data are
res constant
ng the operati
en the adjuste
s, despite the
n-time decisio
scussed issue
field, RMC d
al solutions [
dispatches ar
manage the
. In this pape
wledge from t

han 67 millio
hat is selected
ables is a chal
and Dynamic
s are very u
d that problem
6, 10, 11]. An
n the model. F
, such as a de
model and a m
eeds a major a
dynamically
t review an
ion. If the am
ed model mus
e fact that a
ons.
es and in des
dispatching su
[8, 12]. Cons
re handled by
tasks that are
er, the autho
the decisions
1541
on variables.
d to solve this
llenge.
Data
uncertain and
m conditions
ny change in
For example,
lay of a truck
major change,
adjustment in
changing, so
nd dynamic
mount change
st be reruned.
a dispatching
pite of much
uffers from a
sequently, in
y experts and
e assigned to
rs attempt to
of an expert
.
s
d
s
n
,
k,
,
n
o
c
e
.
g
h
a
n
d
o
o
t
Reconstruction of an Experts Decision Making Expertise in
Concrete Dispatching by Machine Learning

1542
dispatcher. Many types of decisions are made in a day,
however, the authors focus on customer prioritization.
2.3 Role of Experts in RMC Dispatching
RMC dispatchers are expected to satisfy customer
orders, so they must find ways to deliver the
ordered concrete during the day with minimum delay
and cancelations. In reality, there are many parameters
that affect the decisions that an expert takes (Fig. 2).
It should be emphasized that a dispatcher must
simultaneously consider all of these parameters in
reaching a profitable decision. Moreover, prioritizing
customers is one of the most important decisions that
the dispatchers make to achieve a profit. Since the
performance of expert dispatchers is acceptable to
RMCs and the nature of their jobs does not change
significantly, it should be possible to construct rules
that model their decision making.
3. Methodology
Machine learning is used to extract the logic of an
experts decisions from traces of his behavior. In the
present experiments, decision-trees and rule-based
representations were selected for their simplicity and
readability.
Decision tree is have been widely used in machine
learning and data mining [13]. As the authors use the
Weka machine learning toolkit, the J48 algorithms was
selected. This is a reimplementation of Quinlans C4.5
[14]. It uses a divide-and-conquer approach to build
a tree structure that consists of internal nodes, branches
and leaf nodes. An internal node tests a particular
attribute and based on the possible values for the
attribute, a branch is selected. A leaf node represents a
class, so when an instance reaches a leaf node, the
leafs class is assigned to the instance. Decision trees
can handle noisy data by pruning parts of the tree that
may be due to over fitting [14].
The part algorithm [15] creates a set of rules, from
partial decision trees with the aim of avoiding the two
pass method of C4.5 for creating an optimal set of rules
[16].
4. Case Study
To test the proposed method, a simulation model
was built that consists of a batch plant and three

Fig. 2 Summary of affective parameters in RMC (ready mixed concrete).

1543

customers. The simulation model randomly generated
the customers attributes for 200 days. The generated
values for each attribute are within realistic ranges. A
dispatch manager was asked to look at the available
data for all customers and make a decision about the
priority level of each customer. The training sets for
machine learning were constructed from traces of the
dispatchers decision making behaviour. A training
example includes following parameters:
(1) day of delivery (Saturday-Friday);
(2) the amount of ordered concrete for each project
(m
3
);
(3) travel time (min);
(4) spacing time (min);
(5) location of the projects (name of suburbs);
(6) the priority of customers.
Optimization techniques only accept numerical
values and in the objective function all the units must
be the same. For machine learning, the training sets
include a combination of nominal and numerical
variables. Also the selected machine learning
techniques are capable of accepting numerical
attributes with different units.
The number of projects per day in a simulation
model can be varied, however, a simple RMC case was
selected to help the researchers understand the
behaviour of the machine learning techniques and to
give them some insight into the structure of the
database. In each instance, the dispatcher was asked to
simultaneously consider the locations of projects, the
day of delivery in the week and the specifications of all
orders in order to make a decision about the priority of
projects.
To make the simulation realistic, an actual
metropolitan area consisting of seven suburbs was
selected. In this area, there is a batch plant that supplies
concrete for all seven suburbs. The data by the
simulation generated was sent to the dispatch manager
of that batch plant for him to prioritize the projects in
each day. 200 instances were prioritized by the
dispatcher in two stages with each stage involving 100
instances.
In building the training sets, a class must be assigned
to each instance. The machine learning algorithms are
expected to find a relation between the attribute values
of an instance and the assigned classes. The expert
prioritized the customers by providing a string for each
set. For example, if the response of the expert for an
instance is 3-2-1 this means that Customer#3 has the
highest priority which is followed by Customer#2 and
then Customer#1. On the other hand, the number of
training sets is not enough for considering all possible
classes in a learning process. So for each customer a
separate learning process was taken into the account.
So, the assigned level of priority was defined as a class.
For example with reference to the above case the
decision class for Customer#1 is Third, Customer#2 is
Second and Customer#3 is First. Consequently, the
number of classes was dropped from 6 to 3. It provides
more training sets for each class and more accurate
results. However, the training process was taken
separately for each customer. The achieved results for
all customers are very close to others, so for avoiding
redundancy only the results for customer#1 are
discussed.
In the training process, the same dataset was used in
the two selected algorithms, and the 10 folds
cross-validation was selected for evaluating the picked
algorithms. It means that the dataset was divided into
10 folds with around 90% of each fold used for training
and the remaining 10% of data was used for testing.
Then in each fold the performance of algorithms was
calculated for 11 performance metrics, which is further
explained below. Table 1 shows the average of
achieved results from 10 folds for each feature. The
performance metrics used for comparing the
performance of algorithms are: ACC (accuracy), FSC
(F-score), APR (average precision), precision/recall
BEP (break-evenpoint), RMSE (root mean squared
error), CVMS (cross-validation mean sensitivity),

1544
MSP (mean specificity), AUC (area under the ROC),
SAR = (ACC + AUC + (1 RMSE)) / 3, model
building time and model testing time.
In terms of comparison, the most important feature is
ACC. It reflects the ability of each algorithm in
identifying the correct decisions made by a classifier.
According to the achieved results, in terms of accuracy
the performance of J48 with a trivial difference is better
than part. Instances with algorithms that are slightly
different due to randomness issues can not identify
which algorithm outperforms others just based on
accuracy [17-19]. Moreover, the best algorithm in one
feature is definitely not the best in other performance
metrics. In similar cases, the significance test is
recommended. The t-test is a common significance test
for machine learning techniques, which is used for
comparing machine learning algorithm in one domain
[13, 20]. The t-test investigates the meaningful
difference between a pair of algorithms. There are two
hypotheses in a t-test, H0 for equality and H1 for
inequality at an level of significance. The value
is calculated through a t-test and if it is lower than
then H1 will be accepted; otherwise the H0 is true
which reflects that there is no significant difference
between two algorithms. 10 fold cross-validations are
used to calculate the error rate for each fold, and then
the 10 calculated error values for each algorithm were
used in a paired t-test. The derived result of the paired
t-test shows that the P value (0.8088) is greater than
0.01, so can not be proven that there is a significant
difference between the two algorithms. It means that
from a machine learning perspective a 99% level of

Table 1 Results of selected algorithms.
ACC FSC APR PRC RMSE CVMS MSP AUC SAR Building time (s) Testing time (s)
J48 78.921 0.785 0.798 0.798 0.356 0.795 0.104 0.848 0.761 0.0015 0.0000
PART 79.395 0.785 0.797 0.797 0.351 0.802 0.112 0.876 0.773 0.0109 0.0000

Fig. 3 Decision tree produced by J48.
Amount
Customer2
Amount
Customer1
Amount
Customer3
Travel Time
Customer 2
Second
Third First
Amount
Customer3
Third
Amount
Customer 2
Second
Amount
Customer 3
Second
Third
Amount
Custome 3
Amount
Customer 1
Third
Spcaing
Customer2
First
Third
Amount
Customer1
First
Suburb
Customer2
Potts Point Third
Moscot Third
Enmore Second
Alexandria Third
Darlinghurst First
Glebe First
kingsford Third

1545

Fig. 4 Derived rules by PART.

confidence implies that the performances of both of the
selected algorithms are the same. Consequently, the
authors accept the performance of both algorithms and
the built structures by J48 and PART which are
respectively illustrated in Figs. 3 and 4.
At first glance at the results, one can realize that the
most important parameter that the expert dispatcher
considered is the amount of ordered concrete as shown
in Figs. 3 and 4 as under the parameter name Amount
Customer i. The results point out that the higher
priorities were mostly given to larger orders. The next
interesting point that is observed from the results is that
the priority of a customer is assessed not just according
to its own features but also by considering the features
of other customers. This point supports the real
situations/real nature of the experts job and reflects
how the selected algorithm solved the problem.
Moreover, timing attributes and especially the travel
time between the batch plant and the customer were
expected to be affective parameters in the decision
making process. However, the results show that these
parameters are not very important for dispatchers and
that the information about the location of customers
plays a key role in the decisions of an expert. In other
words, the dispatcher mostly took into account the
location of customers with hardly any consideration to
the timing parameters. The expert dispatcher
practically realized the distances between the
customers and the batch plant and the other related
information such as travel time in making decisions.
Moreover, the other important feature that is embedded
in location is the traffic pattern. Some suburbs are
located in crowded areas where traffic is very dense in
rush hours. Therefore, information about the location
Rule 1: IF Amount Customer 2 97 and Amount Customer 3> 113
and Amount Customer1 158 then Third

Rule 2: IF Amount Customer 3 > 195 and Spacing Customer 3 47
and P1_Suburb = Alexandria then Third

Rule 3: IF Amount Customer 3 > 192 and Spacing Customer 3 47
and P1_Suburb = Kingsford then Third

Rule 4: IF Amount Customer1 76 and Amount Customer 2 > 72
and Amount Customer3 90 then Second

Rule 5: IF Amount Customer 2 136 and Amount Customer 1 > 57
and Spacing Customer 1 52 and Amount Customer 3 88 then First

Rule 6: IF Amount Customer 1 > 176 and P2_Suburb = Mascot
and Amount Customer 3 >30 then First

Rule 7: IF Amount Customer 2 > 136 and Amount Customer 3 195
and Amount Customer 1 195 and P1_Suburb = Enmore then Second

Rule 8: IF Amount Customer2 > 163 and Amount Customer1 195
and Amount Customer 3 192 then Second

Rule 9: IF Amount Customer 1 > 176 and P2_Suburb = Kingsford then First

Rule 10: IF Amount Customer 3> 195 and Amount Customer 2 208 then Third

Rule 11: IF Amount Customer 1 107 and Amount Customer 2 134 then Third

Rule 12: IF Amount Customer2 105 then First

Rule 13: IF Spacing Customer1 38 and Amount Customer1 > 62 then First

1546
of a customer provides a lot of information to an expert.
In the conducted case study, sometimes the dispatcher
preferred to postpone a large order by giving a lower
level of priority in order to supply the concrete faster
and more efficiently. In summary, the structures of the
built decision tree and rules are totally different,
however, both algorithms depict a very similar logic in
the humans decisions.
6. Conclusions
In this paper, the extraction of the logic from an
experts decision was conducted in order to build a
system that will be able to automatically handle the
concrete dispatching. To test the proposed method, a
simulation model with a batch plant and three projects
in each day was built. The simulated scenarios were
given to an expert dispatcher who was asked to
prioritize the customers. Then based on the simulation
data and the experts decisions a training process was
conducted on two machine learning algorithms;
decision tree (J 48) and rules-based (PART). In terms of
accuracy both algorithms obtained similar results.
Lastly, though the structures of the decision tree and
rules based are totally different, the derived logic from
both algorithms was very similar. It shows the ways
that experts deal with a complex system such as
concrete delivery problem can be discovered by
machine learning techniques. In the future of research,
the logic extract from large scale RMC dispatching will
be conducted by generalizing problem in machine
learning techniques.
Acknowledgments
NICTA is funded by the Australian Department of
Communications and the Australian Research Council
through the ICT Centre of Excellence program.
References
[1] P.C. Atcin, Cements of yesterday and today: Concrete of
tomorrow, Cement and Concrete Research 30 (2000)
1349-1359.
[2] L. Asbach, U. Dorndorf, E. Pesch, Analysis, modeling and
solution of the concrete delivery problem, European
J ournal of Operational Research 193 (2009) 820-835.
[3] P.C. Lin, J . Wang, S.H. Huang, Y.T. Wang, Dispatching
ready mixed concrete trucks under demand postponement
and weight limit regulation, Automation in Construction
19 (2010) 798-807.
[4] S. Yan,W. Lai, An optimal scheduling model for ready
mixed concrete supply with overtime considerations,
Automation in Construction 16 (2007) 734-744.
[5] C.W. Feng, T.M. Cheng, H.T. Wu, Optimizing the
schedule of dispatching RMC trucks through genetic
algorithms, Automation in Construction 13 (2004)
327-340.
[6] S. Yan, W. Lai, M. Chen, Production scheduling and truck
dispatching of ready mixed concrete, Transportation
Research Part E: Logistics and Transportation Review 44
(2008) 164-179.
[7] M. Maghrebi, T. Waller, C. Sammut, Scheduling concrete
deliveryproblems by a robust meta heuristic method, in:
Sevens European Symposium on Computer Modeling and
Simulation (EMS), UK, 2013.
[8] D. Naso, M. Surico, B. Turchiano, U. Kaymak, Genetic
algorithms for supply-chain scheduling: A case study in
the distribution of ready-mixed concrete, European
J ournal of Operational Research 177 (2007) 2069-2099.
[9] V. Schmid, K.F. Doerner, R.F. Hartl, M.W.P. Savelsbergh,
W. Stoecher, A hybrid solution approach for ready-mixed
concrete delivery, Transportation Science 43 (2009)
70-85.
[10] V. Schmid, K.F. Doerner, R.F. Hartl, J .J .
Salazar-Gonzlez, Hybridization of very large
neighborhood search for ready-mixed concrete delivery
problems, Computers & Operations Research, Operations
Research 37 (2010) 559-574.
[11] Z. Yang, L. Menglei, L. Zhenyuan, Vehicle scheduling
and dispatching of ready mixed concrete, in: Fourth
International Workshop on Advanced Computational
Intelligence (IWACI), China, 2011, pp. 465-472.
[12] C.W. Feng, H.T. Wu, Integrating FMGA and CYCLONE
to optimize the schedule of dispatching RMC trucks,
Automation in Construction 15 (2006) 186-199.
[13] I.H. Witten, E. Frank, Data Mining, Practical Machine
Learning Tolls and Techniques, Elsevier, New York,
2005.
[14] R. Quinlan,. C4.5: Programs for Machine Learning,
Morgan Kaufmann Publishers, San Mateo, 1992.
[15] E. Frank, I.H. Witten, Generating accurate rule sets
without global optimization, in: Proceedings of the
Fifteenth International Conference on Machine
Learning, Morgan Kaufmann Publishers Inc., 1998, pp.
144-151.
[16] W. Buntine, T. Niblett, A further comparison of splitting

1547
rules for decision-tree induction, Machine Learning 8
(1992) 75-85.
[17] E. Rahm, P.A. Bernstein, A survey of approaches to
automatic schema matching, The VLDB Journal 10 (2001)
334-350.
[18] T.G. Dietterich, An experimental comparison of three
methods for constructing ensembles of decision trees:
Bagging, boosting and randomization, Machine Learning
40 (2000) 139-157.
[19] S. Garcia, A. Fernandez, J. Luengo, F. Herrera, A study of
statistical techniques and performance measures for
genetics-based machine learning: Accuracy and
interpretability, Soft Computing 13 (2009) 959-977.
[20] T.G. Dietterich, Approximate statistical tests for
comparing supervised classification learning algorithms,
Neural Computation 10 (1998) 1895-1923.


Medium-Deep or Very Deep Disposal of Highly
Radioactive Waste?
Roland Pusch
1
, Gunnar Ramqvist
2
, Sven Knutsson
1
and Mohammed Hatem Mohammed
1

1. Department of Civil, Environmental and Natural Resources Engineering, Lule University of Technology, Lule SE-971 87,
Sweden
2. Eltekno AB, Figeholm SE-57295, Sweden

Abstract: Commonly proposed concepts, like KBS-3V, for disposal of highly radioactive waste imply construction at medium depth
(400-600 m) in granitic rock, which is excellent for constructing a stable repository. VDH (very deep boreholes) represent an
alternative concept with the advantage that the rock is much less permeable and that the very salt, heavy groundwater is stagnant. Both
require engineered barriers in the form of canisters and waste-embedding clay but for somewhat different purposes. Canisters are the
most important waste-isolating barriers for KBS-3V but are less important for VDH. The waste-embedding clay is needed for
preserving the KBS-3V canisters by being tight and ductile, but plays a minor role for the VDH. The backfilled deposition tunnels in a
KBS-3V repository provide very limited hindrance of radionuclides to move to the biosphere while the clay seals of VDH effectively
prevent possibly released radionuclides to reach up to the biosphere. Comparison of the KBS-3V and VDH concepts indicates that the
last mentioned one has several advantages but that certain issues remain to be worked on for becoming a number one candidate.

Key words: Canisters, deep boreholes, groundwater, radioactive waste, site selection, super containers, waste disposal.

1. Introduction

Several comprehensive studies like the CROP
project [1] have been conducted for assessing the
capacity of proposed repository concepts for isolating
HLW (highly radioactive waste) from the biosphere.
Focus was on crystalline rock that requires more
effective engineered barriers than the others because
of its higher permeability but provides more stable
conditions and is the main candidate for hosting
repositories in Sweden, Finland and Canada. URLs
(underground laboratories) for testing the performance
of rock and EBS (engineered barriers) have shown
that the variability of the properties of the rock makes
prediction of its role difficult but that the function of
EBS is predictable. The URLs represent a large
enough variety of conditions for providing credible
generic information on repository concepts. The
present document summarizes and compares the main

Corresponding author: Roland Pusch, M.Sc., Ph.D.,
professor, research field: soil mechanics. E-mail:
drawrite.se@gmail.com.
functions of the host rock and EBS of repositories for
disposal of HLW at medium and large depths.
Commonly proposed depths for locating a
repository of one-level type is 400-600 m with access
to the repositories through ramps or shafts. As to the
host rock the structure is important because it controls
the groundwater flow and the impact on the disposal
site by earthquakes and tectonic movements. The most
important EBS component is the metal canisters
containing the waste (spent fuel or vitrified, processed
waste), and the buffer clay surrounding them for
minimizing groundwater flow in their vicinity and for
providing ductile embedment (Fig. 1). The waste
generates a heat pulse that lasts for several hundred
years and can strongly affect the function of the
barriers and the groundwater flow, as well as the stress
conditions in the rock, and thereby its physical
stability.
VDH (deep hole concepts) imply waste placement
in the lower part of 3-5 km deep holes and effective
sealing of the upper part (Fig. 2) [2-5]. Several holes
DAVID PUBLISHING
D
Medium-Deep or Very Deep Disposal of Highly Radioactive Waste?

1549
Assemblyofwell
fittedblocksof
highlycompacted
smectitic clay
Densepelletsof
smectitic clay
blowntofillthe
spacebetween
blocksandrock
Blocksofhighly
compactedblocks
ofsmectitic clay
HLWcanister
Boreddeposition
holeswith68m
spacing
Bottombedof
concretecaston
site
Blastedtunnel

Fig. 1 Schematic section of a tunnel with deposition holes
with clay-embedded canisters according to the KBS-3V
concept [1].

will be bored from a shallow chamber so that the
distance between the waste-bearing parts is
sufficiently large to avoid interference and
superposition of temperature fields, and sufficiently
small to create a common space for rational boring. A
recent version, discussed in this paper, has a constant
diameter of 0.8 m. Stability of the holes requires that
they are mud-filled and provided with casings over the
whole borehole length except where concrete seals are
constructed.
2. Host rock
2.1 Structure
For technical/scientific purposes and for predicting
the thermal, mechanical and hydraulic performances
of the host rock in bulk one needs to use definitions of
discontinuity elements with typical properties
respecting the hydraulic function and rheological
performance. Major, water-bearing structural elements,
represented by dominant fracture zones termed 1st and
2nd order discontinuities [6, 7], should not intersect
waste-containing parts of a repository and are hence of
primary importance in the site-selection process
(Fig. 3). They can be identified by examination of
cores from deep, vertical and graded boreholes and by
in-situ logging using geophysical methods. Finer ones
are 3rd order discontinuities, i.e., minor fracture zones,
and those of 4th order being discrete, persistent
fractures, determine the performance of the rock
where waste containers are placed. They remain
unidentified until construction work has started. Even
finer discontinuities, i.e., those of 6th and 7th orders,
belong to the rock matrix and represent its content of
pores and fine fissures.

0.5km
1.5km
2.0km
Bottomat4kmdepth
SetsofHLWcanisters
separatedbyblocksof
highlycompacted
smectiterichclay
Setsofblocksof
highlycompacted
smectiterichclay
Uppersealofconcreteundermetal
plug
Drawrite AB
0.8m

Fig. 2 Schematic view of a VDH concrete to 0.5 km depth over tightly sealed part to 2 km depth, the disposal zone with
HLW canisters reaches to 4 km depth [5].

1550

Fig. 3 Simplified model of a repository site derived from core mapping and geophysical measurements, light green area is the
ground surface and blue plates 2nd order discontinuities, red plates are of 3rd order black panels are HLW deposition tunnels
[7].

Fig. 4 shows a structural model derived on the basis
of core mapping and geophysical investigations
including in-situ hydraulic measurements. The
discontinuities represent fracture zones of 2nd order
forming distorted and curved, more or less orthogonal
patterns. Fig. 5 illustrates the pattern of actually
identified 4th order fractures at 360 m depth in
another area, showing similar winding nature of
discontinuities also on this smaller scale. The
orientation of the major principal plane follows
approximately the orientation of one of the major sets,
the highest horizontal pressure ranging between 20
MPa and 28 MPa.
2.2 The Waste-Isolating Role of the Host Rock
2.2.1 Major Criteria
Safe disposal of HLW requires that there is nearly
no dissemination of possibly released radionuclides in
at least 100,000 years as stipulated by Swedish
authorities and those of several other countries. While
the function and role of engineered barriers, primarily
the waste canisters and their clay embedment, are well
defined and predictable, the properties and function of
rock hosting repositories at a few hundred meter depth
vary so much that its role as barrier is being
questioned. In recent time, it has been degraded to be
a mechanical protection of the chemical apparatus
[5]. The major reason for this is the difficulty in
working out reliable models of its hydraulic
performance, especially in a long-term perspective,
because of the very limited information on the
structure, which is a consequence of the winding and
unknown interconnectivity and persistence of both
low- and high-order discontinuities. A second reason
is that the predicted degree of utilization, expressed as
the number of acceptable positions of HLW containers
per unit length of a KBS-3 deposition tunnel or per
meter length of the deployment segments of VDHs,
depends on the initially unknown frequency of
intersected 2nd and 3rd order discontinuities and of
the stability of the rock around the deposition holes.
The role of the host rock is fundamentally different
for the two concepts KBS-3V, taken as a
representative of medium-deep repositories and VDH.
For KBS-3V, it is to provide mechanical protection of
the waste while for VDH, it is to prevent groundwater

1551

Fig. 4 Perspective view of 16 km
3
rock volume with 33 steep 2nd order fracture zones intersected by 1 km deep holes: 19
zones by KLX 04, 4 zones by KLX 06, and 10 zones by KLX 10. Spheres indicate the upper ends of the holes [8].

Fig. 5 Winding of structures and orientation of the major horizontal stress (fat arrows) in system of 4th order discontinuities
at 360 m depth, the spacing of the lines is 30 m [9].
A
B
C
D
Meas. stress on
principal planes
A 25 MPa
B 28
C 24
D 20
MPa
MPa
MPa
MPa


1552

at the waste level to move up to shallow, permeable
rock. This is fulfilled by the high salt content that
makes possibly contaminated groundwater stay at
depth.
2.2.2 Utilization
(1) KBS-3V
For a KBS-3V repository, the possibility to locate
deposition holes, which are 1.95 m in diameter and
about 8 deep, can not be predicted until rock
excavation has reached the planned repository level
and the position and orientation of transport and
deposition tunnels have been decided [1]. After
constructing these tunnels, which will provide rich
information on the rock structure and probably cause
changes of the plans, slim pilot holes will be drilled in
different directions for general exploration and in the
direction of tunnel positions proposed on the basis of
gained information on structural features. A few holes
with 3-400 m length drilled within the 20 m
2
section
of forthcoming tunnels will reveal the presence and
orientation of 3rd order zones and give a first
indication of whether the plan is worth pursuing. A
large number of deposition hole positions must be
investigated for making sure that none of them will be
intersected by other than 4th and higher order
discontinuities before the entire repository design can
be decided. Preparative work comprises boring of
pilot holes for determining the uniaxial compressive
strength and the shear strength of intersected fractures
of 4th order type for predicting the degree of
mechanical stability of the rock.
The site investigation and design work will take at
least a decade for a repository with 6,000-7,000
canisters but will ultimately provide plans of the type
shown in generalized form in Fig. 6. They indicate the
degree of utilization, which is commonly 50%-60%,
representing the percentage of useful hole positions.
Full-face boring of deposition holes will be made by
using TBM-type technique;
(2) VDH
For a VDH repository, the basis of evaluation of the
degree of utilization for waste placement is much
simpler since it is confined to identifying 2nd and 3rd
order discontinuities that will intersect the deep holes.

Fig. 6 Schematic plan view of KBS-3V repository with transport and deposition tunnels in rock with typical frequency of 3rd
order fracture zones (D3), tight bulkheads are keyed into the rock for isolating deposition tunnels [7].
50-100 m
30 m
30-40 m
N
Deposition hole
PERMPLUG
with clay in key
PERMPLUG
without clay
D3 Fracture zone
D3 Fracture zone
D3 Fracture
zone
D3 Fracture zone Deposition tunnels
Transport tunnel
Wide EDZ
Normal EDZ
TEMPLUG
TUNNFILL
QUARTZ SAND

1553

This only requires a prospective investigation by slim
core drilling, rock characterization and hydraulic and
strength testing for deciding how frequently
stabilization and placement of waste canisters shall be
made. Subsequent to this, detailed design and boring
of a full-size VDH can start. The matter of where
excavated rock shall be transported and stored is an
important environmental issue for mined KBS-3V
repositories but is of no concern for VDH.
An example of what the conditions are with respect
to the degree of utilization in a 1,000 m deep hole at
the HLW disposal site proposed by the Swedish Fuel
and Waste Handling Co (SKB) is illustrated in Table 1.
The example indicates that the parts that can be used
for waste placement represent about 50%. There are,
however, strong indications that the spacing of
water-bearing fracture-rich zones is larger at depth as
indicated by the Gravberg-1 boring in crystalline rock
[4], which showed extensive fracturing down to
1,200 m depth but only 2 m to 20 m fracture zones
with a spacing of at least 200 m deeper down. This
would imply a percentage of 90% or more. A cautious
estimate would be 75%.
2.3 Constructability and Short Term Performance
A major criterion for accepting and realization of a
HLW repository is that it is constructable. This has
been demonstrated by the URLs representing
medium-deep disposal according to KBS-3V like
concepts but not yet for VDH. For the rock, it must be
shown that the strength is sufficient for preventing
initiation of practically important fracturing and fall of
rock fragments from the borehole walls.
2.3.1 KBS-3V
For the KBS-3V deposition holes with 6 m spacing,
1.95 m diameter and 8 m depth in a repository at
400-500 m depth in rock with the typical horizontal
rock stresses 20 MPa and 40 MPa and vertical stress
10 MPa, the theoretical maximum hoop (tangential)
stress is about 120 MPa assuming the least critical
orientation of the tunnel [10].
Using the unconfined compressive strength, which
is commonly 150-350 MPa for granitic rock, as
practical parameter for assessing hole positions,
application of statistical methods can be made for
finding the degree of utilization. For this purpose, one
needs to apply a lower value of the compressive
strength since initiation of macroscopic failure in fact
takes place at about 50% of the conventionally
determined strength value [11]. Adopting this
principle it has been found that a few percent of the
positions have to be abandoned in the site selection
process as exemplified by the ongoing construction of
SKBs HLW repository at Forsmark in Sweden. The

Table 1 Major structural features in 1,000 m deep borehole KFM07A [8].
Depth (m) Structural features Useful for placing waste canisters
0-210 Major 2nd and 3rd order fracture zones (DZ1, DZ2) Unsuitable
210-260 Fracture-poor Suitable
260-290 Fracture-rich Unsuitable
415-435 Major 2nd order fracture zone (DZ3) Unsuitable
535-545 Fractures Acceptable
790-1,000 Major 2nd order fracture zone (DZ3) Unsuitable


1554

winding of fractures, fracture zones and major
horizontal stress, means that a certain fraction of the
deposition tunnels will be less favorably oriented in
the stress field and, that some additional positions
have to be abandoned. The majority of the planned
holes would still be acceptable taking also the risk of
activation and propagation of natural, partly hidden
discontinuities of 5th and 6th order discontinuities into
consideration [7]. However, the situation will be
totally different when the heat pulse emerges from the
radioactive decay. The hoop stress will then rise and
spalling occur in most of the deposition holes, causing
a very substantial increase in vertical hydraulic
conductivity along them [10].
Rock fall into an empty KBS-3V hole in the
canister-placement phase can be very problematic
since a wedge of any size coming off from the rock
wall can hinder further deposition or upward pulling
of the canister (Fig. 7).
2.3.2 VDH (Very Deep Boreholes)
For VDH, the matter of stability is different because
of the higher stresses. Using data bases and
generalized relationships between depth and stress
states [6] the maximum and minimum horizontal
stresses can be estimated at 56 MPa and 39 MPa at 2
km depth, but higher pressures have been recorded at
this depthas in the 6 km deep German KTB hole [12].
Here, the horizontal stresses were found to be 70 MPa
and 30 MPa at 2 km depth, and 120 MPa and 70 MPa
at 4 km depth. High pressures have also been reported
from measurements at 3 km depth in crystalline rock
as in the Nojima fault in Japan where the horizontal
stresses were 88 MPa and 64 MPa, respectively [13].
Assuming, for a forthcoming VDH in Sweden, the
same stresses as in the German KTB hole, the hoop
stress around a bored hole would be up to 100 MPa at
2 km depth and 190 MPa at 4 km depth. No correction
for the impact of winding stress directions would be
necessary below 2 km since the variation in
orientation of principal stresses is reported to vary
only within 10
o
[14].

Fig. 7 Wedge formed by intersection of critically oriented
4th order discontinuities.

As for the KBS-3V, the thermal stresses will add to
the static pressure in the deployment part of a VDH
but to a lesser extent because of the much smaller
temperature rise. At 4 km depth, the ambient
temperature is expected to be about 64
o
C [4] and for
two possible canister configurations, one with 4 BWR
spent fuel elements and one with 2 BWR and 1 PWR,
the net temperature will be 90
o
C at 2 km depth and
150
o
C at 4 km depth assuming common data of the
heat generation and thermal properties of the EBS and
the rock [3, 4].
2.4 Waste Placeability
The installation of clay blocks according to the
KBS-3V implies individual placement to give a
straight column of annual blocks in which a canister is
inserted remotely by robot technique (Fig. 8). The
procedure causes a risk for getting the canisters stuck at
lowering them into the tight space, the tolerance is only
10 mm.
For VDH the placeability is determined by the
straightness of the holes. If the waste packages are too
long they can become stuck in curved boreholes [8].
For a borehole diameter of 800 mm and an inner
diameter of the rock-supporting casing of 750 mm,
super containers with 700 mm outer diameter and 25
m length can be moved down in the clay mud without
significant difficulties. Fig. 9 illustrates the variation
in curvature of slim boreholes, which is believed to be

1555

(a) (b)
Fig. 8 Placement of clay blocks and canisters in canister deposition holes in a KBS-3V repository: (a) buffer block with 1.85
m diameter and 0.4 m height weighing about 2,000 kg for installation in vertical deposition hole; (b) canister installation [1, 7].

valid also for a full-size VDH using a slim hole for
guidance.
All super container units rest on previously inserted
sets except where fracture zones are intersected. Here,
concrete is cast, after local reaming of the hole, up to
the level where the fracture zone ends. The next sets
of canister-containers are inserted when the concrete
has matured sufficiently for getting the required
bearing capacity. Schematically, a VDH unit for
placement in the deployment zone would look as in
Fig. 10. In the overlying sealed zone, with no waste
the super containers look the same but contain only
dense clay blocks. In the deployment zone, super
containers filled with only clay blocks can be
incorporated between those with waste for avoiding
criticality.
3. Engineered Barriers
3.1 Types and Roles
Multi-barrier concepts like KBS-3V and VDH rely
on the rock as natural barrier and on man-made
barriers: canisters with HLW and clay surrounding
them, and clay and concrete for sealing shafts and
holes that lead from the ground surface to the waste.
The temperature, to which the barriers will be exposed,
can be up to 100
o
C for KBS-3V and 150
o
C for VDH,
and change their waste-isolating capacity. This
requires assessment of their role in the respective
repository environment.
3.2 Canisters
For the KBS-3V and VDH concepts metal waste
canisters serve as primary engineered barriers. The
presently favored KBS-3V canister consists of cast
iron with 12 BWR elements of spent fuel, lined with
50 mm copper, and the latest VDH canister type,
favored by the present authors, is of the same type but
with only 4 BWRs. The outer diameter of the KBS-3V
canister is 1,050 mm and the height 4,850 mm, its
weight being 24,500 kg. The VDH canister, fitted in
perforated super containers, is proposed to have 500
mm diameter and 4,850 mm height, weighing about
6,000 kg. It can be made of an alloy Navy Bronze,
consisting of more than 90% copper and about 10%
nickel for fulfilling the criterion of mechanical
strength and chemical compatibility [6]. The
rock-stabilizing casing in VDH should be made of the
same material
1
. A small fraction of the canisters will
have defects from start and leak, which contaminates
the surrounding clay and the surrounding rock with
radionuclides. Depending on the degree of ductility
of the clay, thermally or seismically generated

1
The use of titanium for canisters, super containers and casings
is presently being investigated.
Canister being placed
in deposition hole
Blocks of highly compacted
smectite clay
Gap between buffer blocks
and rock filled with smectite
pellets according to SKBs
concepth KBS-3V
EDZ

1556

Fig. 9 Example of logging of cored hole in the Forsmark region [8].

Clay
block
Upperendofsupercontainer
withcanistersandclayblock
Canisters
Casingfor
supporting
borehole
Borehole
0.8m
diameter
Supercontainer
Fracturezone
Fracture
poorrock
Drawrite AB
Concrete

Fig. 10 Super containers in the deployment zone for placement in clay mud [5].

displacements in the rock will be transferred to the
canisters that may ultimately fail [7]. Water will
migrate into the canisters and make the iron core
corrode and produce highly pressurized hydrogen gas
that penetrates the surrounding clay channel wise and
migrate further along the rock contact or within the
rock. The canisters in a KBS-3V repository will be
exposed to tension caused by upward expansion of the
buffer clay [7]. This movement can be very significant
depending on the compressibility of the overlying
tunnel backfill and can cause critical stresses and
breakage at the junction of copper liner and lid. VDH
canisters of SKB type will be under high isostatic
pressure but there will be no tension.
3.3 Clay
Clay serves as a very effective sealing component
in both concepts by retarding groundwater flow
around and along the canisters and by providing a
homogeneous ductile embedment of the canisters [1].
It must be tighter than the surrounding rock and stay
ductile during the entire operational time, which is
KFM07A
100
1000
10000
100000
0 100 200 300 400 500 600 700 800 900 1000
Borehole Length (m)
R
a
d
i
u
s

o
f

c
u
r
v
a
t
u
r
e

(
m
)
100,000
10,000
1,000
100
0 100 200 300 400 500 600 700 800 900 1,000
Borehole length (m)

taken as 100
requires use
montmorillo
palygorskite
acceptable o
concept, th
containers, i
montmorillo
components
useful for tw
compacted b
The matur
the mud and
different and
differently. F
will be re
deposition
saturation b
period, the d
and salt en
processes on
expandabilit
water satura

Fig. 11 Sch
hydrotherma
constant [15].
M
0,000 years i
e of expandab
onite and sap
e and mixed-l
or even favo
he buffer c
is presently p
onite. For th
of the sam
wo purposes
blocks.
ration of the
d dense clay
d affects thei
For the first m
latively low
holes low,
by tens to hu
desiccated cla
nrichment, d
n the micro-s
ty and self-he
ated clay, and
hematic diffra
ally treated sam
.
Medium-Deep
in Sweden a
ble clay of sm
ponite, althou
layer types m
orable [2]. F
lay surround
proposed by S
he VDH con
me type as f
s, i.e., as mu
buffer clay in
components
r long-term c
mentioned, th
w and the i
which ca
undreds of ye
ay is exposed
dissolution a
structural scal
ealing ability
also make it
actograms of
mple (130
o
C)
or Very Deep
and Finland.
mectite type,
ugh minerals
may turn out t
For the KBS
ding the w
SKB to be ric
ncept, two
for KBS-3V
ud and as hig
n KBS-3V an
in VDH is q
chemical stab
he water pres
inflow into
an delay w
ears. During
to at least 10
and precipita
le can reduce
of the ultima
stiffer [6, 15
a montmorill
. Feldspars di
p Disposal of
This
like
like
to be
S-3V
waste
ch in
clay
are
ghly
nd of
quite
bility
ssure
the
water
this
0
o
C
ation
e the
ately
].
In
tem
cau
tigh
part
cha
indi
M
(non
sme
qua
acc
Cem
Nat
indi
exp
a s
pres
be t
T
mon
[20
perf
lonite-rich ref
sappeared in
f Highly Radio
n the uppe
mperature wil
using no heat
htness early
t, which will
anges and le
icated by Fig
Montmorillon
n-expansive)
ectite/illitemi
artz will be pr
ess to pot
mentation is c
tural analogu
icate that T
posed to 130-
significant pa
served, and a
the case for K
The model
ntmorillonite
] and used
formance gi
ference clay (M
the hot sampl
oactive Wast
er, sealed p
ll range from
t-induced cha
after installa
be heated up
ess good iso
g. 11 [16-18].
nite will
illite
nerals and sil
recipitated at
tassium and
caused by ne
ues from var
Tertiary and
-160
o
C for a
art of their
a common be
KBS-3V and V
of mine
proposed b
in various a
ives the ra
MX-80), samp
le but the cont
e?
part of the
m about 15
o
anges and pr
ation. In the
p to 150
o
C, m
olation are

be con
via
licious miner
t a rate determ
d temperatur
eoformed qua
rious parts o
d Ordovician
few thousan
montmorillo
elief is that t
VDH [7].
eralogical c
by Grindrod
assessments o
ate of disso

ple, untreated
tent of montm
1557
e VDH the
o
C to 60
o
C
oviding total
deployment
mineralogical
expected as
nverted to
mixed-layer
rals including
mined by the
re [7, 19].
artz and illite.
of the world
n bentonites
nd years have
onite content
this will also
changes in
and Takase
of repository
olution and
d (20
o
C) and
morillonite was
7
e
C
l
t
l
s
o
r
g
e
.
.
d
s
e
t
o
n
e
y
d
s

1558

precipitation of phyllosilicates by taking O
10
(OH)
2
as
a basic unit and defines a general formula for smectite
(S) and illite (I) as:
X
0.35
Mg
0.33
Al
1.65
Si
4
O
10
(OH)
2
and K
0.5-0.75
Al
2.5-2.75

Si
3.25-3.5
O
10
(OH)
2
(1)
where, X is the interlamellar absorbed cation (Na) for
Na montmorillonite. According to the model, the rate
of the reaction can be expressed as:
r = Aexp (-E
a
/RT)(K
+
)S
2
(2)
where, A is coefficient, E
a
is activation energy for the
conversion of montmorillonite to illite (S/I), R is
universal gas constant, T is absolute temperature, K
+
is
potassium concentration in the pore water, and S is
specific surface area for reaction.
Applying this model to the two concepts the
following chemical evolution of the clay is foreseen:
(1) KBS-3V
Silica will be released and transported from the
hottest to the coldest part of the clay buffer [6].
Assuming a linear temperature drop with time to 25
o
C
after 10,000 years, silicification and illite formation
would be initiated after 500 years, the first mentioned
within about 0.1 m from the rock, and the latter
occurring in the hotter part (60-100
o
C). Since the pore
water will be saturated with silica in the hot period it
will be precipitated also in the most heated part when it
cools. Thee hydrothermal reaction products silica and
illite will cause cementation and significant reduction
in expandability of the buffer [8];
(2) VDH
The mineralogical stability of the VDH clay
components in the upper part is similar to that of the
KBS-3V buffer, while the mineral content of the mud
and clay in the deployment part will be transformed
and serve differently with time. The high
temperature100
o
C to 150
o
Cis believed to give
essentially the same effects but the rates of creation of
illite and stiffening will be much higher than for
KBS-3V and the colder part of VDH. Thus, raising the
temperature from 100
o
C to 150
o
C speeds up the rate
of illitization by about 100 times, assuming the
activation energy to be 27 kcal/mole and taking all
other factors to be the same. This significant loss of
effective sealing will take place in about 100 years but
the practical importance is small since the tightness of
the converted clay is still considerable. Thus, for pure
illite with a density of 1,600 kg/m
3
the hydraulic
conductivity is lower than E-8 m/s [21, 22] and a
hundred times lower than this for the density 1,900
kg/m
3
. Total conversion to non-expandable illite in a
dominant part of the dense clay in the supercontainers
would in fact take 100,000 years because the
controlling mechanism is diffusive transport of
potassium from the surrounding rock [7].
Exposure of water saturated montmorillonite-rich
clay to strong gamma radiation and 90-135
o
C
temperature has no degrading impact according to
French investigations but speeds up the release of Fe
from metal iron like iron or steel canisters [15].
3.4 Concrete
Concrete will not be used in a KBS-3V repository. In
a VDH concrete seals will be constructed where it
intersects major discontinuities like fracture zones of
2nd and 3rd orders [8]. The main criteria for the
concrete seals in VDH are: (1) sufficient fluidity
forcasting; (2) sufficient bearing capacity for carrying
the overlying super containers; (3) lower hydraulic
conductivity than of the fracture zone to be sealed; (4)
insignificant chemical impact on contacting clay seals.
They are fulfilled by using talc-based concrete with a
density of about 2,000 kg/m
3
and a low content of very
fine-grained low-pH cement (< 8 weight percent of the
solids) [23].
4. Comparison of SKB-3V and VDH
Functions
4.1 Rock Stability
The essential difference between the two concepts is
that the stability of the deep vertical VDH with 800 mm
diameter becomes acceptable by using a dense
deployment mud, serving also as drilling mud, while

1559
that of KBS-3 holes is insufficient after placement of
the canisters and can not be improved by technical
means. The impact on the maximum radial stress by the
mud is demonstrated by inserting the depth-dependent
mud pressure in the classical Kirsch equations [6],
leading to the expression in Eq. (3) for the hoop stress:
= (
H
+
h
) 2(
H

h
) cosgz (3)
where,
H
and
h
are the major and minor horizontal
stresses, is the angle between the considered plane
and the plane in which the major horizontal stress is
located, is the density of the mud, g is the coefficient
of gravity and z is the depth.
For the same rock properties as those assumed for
KBS-3V and taking the uniaxial compressive strength
according to SKB, i.e., 150-350 MPa, the rock in the
deepest part of the deployment zone will be stable by
using mud is present. As discussed earlier here, one
can assume the same rock stresses as in the German
KTB case, implying the hoop stress around a bored
hole to be up to 100 MPa at 2 km depth and 190 MPa
at 4 km depth. With only water in the hole, the hoop
stress would be 208 MPa at the bottom of the 4 km
deep holes but with mud weighing 1,600 kg/m
3
the
hoop stress would be reduced to 168 MPa and to 85
MPa at 2 km depth. One hence finds that boring of a 4
km deep VDH using even very dilute clay mud would
be possible without risk of spalling and fall of rock
wedges. For the proposed denser, mud stability is
definitely provided but a supporting casing indicated in
Fig. 10 is recommended for extra support.
The heat production of the HLW in a KBS-3V
repository extending over of a square kilometre and
located at 450 m depth has been found to give an
accumulated maximum rise of the ground surface of a
couple of decimetres after 1,600 years [24] and
significant shearing of discontinuities of 2nd and 3rd
orders. This can affect the hydraulic performance of the
repository rock and make predictions thereof very
uncertain. For a VDH repository, such impact is
negligible since the heat sources, i.e., the waste in the
deployment parts of a group of VDH holes at the
respective site, will be at least 200 m apart.
A possible impact of high deviatoric stresses is
propagation of subhorizontal natural weaknesses of
2nd or 3rd orders crossing a KBS-3V repository
(Fig. 12). It can be triggered by a thermally raised pore
pressure in the deposition tunnels causing reduction of
the effective vertical rock stress by which further slip is
facilitated leading to simultaneous shearing of many
canister deposition holes.
4.2 Physical Evolution of EBS Clays
In contrast to KBS-3V all parts of a VDH the clay
blocks are quickly water saturated under the
prevailing high water pressure. The ultimate density of
the clay blocks in both types will be 1,900 to 2,000
kg/m
3
, which gives the clay a hydraulic conductivity
of no more than E-11 m/s at saturation with
Ca-dominated water, and a swelling pressure of at
least 1 MPa [6]. Over the larger part of the length of
the super containers containing HLW canisters the
density of the mud will remain unchanged at 1,600
kg/m
3
(dry density 950 kg/m
3
) but the physical
properties will change, implying an increase in
hydraulic conductivity and a drop in swelling pressure
[15]. They will be affected by the salt concentration in
the groundwater and exchange of the initially sorbed
sodium ions by copper ions and Ca, being the
dominant cation in the strongly brackish KBS-3V

Fig. 12 Large horizontal break formed by propagation of
an embryonic weakness extending from a natural major
discontinuity of 2nd or 3rd order (A) [7].
Propagation of
break
Deposition tunnels or drifts
Transport tunnel
Large horizontal break emanati ng
from maj or natural disconti nuity A
A

1560
groundwater and in the very salt groundwater (> 10
g/L) in the deeper parts of a VDH [6]. They will
control the micro-structural constitution and thereby
the physical properties.
The VDH concept discussed here implies that the
boring mud is the same as the deployment mud. For
serving in the boring phase, it must have the required
ability to support the hole being bored and to bring up
the debris from the boring head. The super containers
with waste canisters and dense blocks of expansive
clay shall sink down or be pushed through the mud to
the predetermined levels, which requires that its shear
resistance is sufficiently low. After placement the mud
will be consolidated by the expanding dense blocks in
the super containers, intermittently sealing off the
holes in the deployment zone, and completely
tightening the holes in the upper, sealed part. For the
deployment part, where the role of the heavy salt water
is most important for preventing possibly released
radionuclides to move above it, the only criterion is that
the consolidated mud should stay homogeneous and in
tight contact with the rock and canisters for minimizing
convective water flow.
For allowing a set of super containers weighing
about 20,000 kilograms to sink or be pressed down in
the mud its shear resistance must be sufficiently low.
The resistance to movement is caused by the point
resistance and the friction at the slip along the super
container. The theoretical point-bearing capacity of
clay is 5-10 times the shear strength
2
[25]. The shear
strength required for carrying a 20,000 kg set of super
containers exerting a downward pressure on the mud of
about 0.5 MPa is hence on the order of 50 kPa to 100
kPa. It can be estimated from the swelling pressure and
the angle of internal friction, which can be taken as 15
o
,
giving the approximate relationship between the
density and shear strength in Table 2. It shows that the
highest acceptable mud density for making the super
container sets sink without adding loads is about 1,600

2
The short period of time of installation means that the
undrained shear strength is the relevant measure of strength.
kg/m
3
. Once failure has been initiated and the super
container has started slipping down, the wall friction
along its outer surface is mobilized and provides
additional resistance. It is the product of the sheared
surface area and the residual shear strength, which, for
smectite clay, can be a fraction of the shear strength at
rest. Depending on the length of the containers they
will slip down without additional force or require
pushing. As for pile driving in such soils a vibrator
with variable frequency may have to be used if the
super container sets do not sink. The downward
movement is associated with upward flow of
remoulded mud along them.
5. Possible Improvement of the Concepts
5.1 General
Both concepts have weaknesses or uncertain
functions and current research and development aims
at elimination of the question marks. The authors will
consider some recently proposed improvements that
are aimed at giving them higher credibility and larger
applicability for no extra cost.
5.2 KBS-3V
The major problem with this concept is the
overstressing caused at the boring of a certain number
of deposition holes and of all the holes when the
thermal pulse is generated. A recently proposed version
(KBS-3i) that minimizes this problem has deposition
holes with every second directed 45
o
off the verticalto
one side, and the others to the opposite as shown by
Fig. 13 [7]. This reduces the hoop stress by about 20%
as illustrated by Fig. 14, which also shows a way of

Table 2 The shear strength of Na smectite-rich
(montmorillonite) clay of different densities.
Density (kg/m
3
) Swelling pressure (kPa) Shear strength (kPa)
1,300 45 12
1,400 100 27
1,500 200 54
1,600 350 95
1,700 550 149
1,800 800 216

1561

Fig. 13 The KBS-3i concept with 45
o
dip and +/-45
o
deviation from the tunnel axis, the clay granulate is tentative [7].

(a) (b)
Fig. 14 Plotting of von Mises stresses in sections through deposition holes normal to the hole axis: (a) KBS-3V; (b) KBS-3i.
Red zones representing critically high stresses are much larger for KBS-3V [10].

Fig. 15 Technique for stabilizing boreholes: (a) borehole intersecting fracture zone; (b) reamed hole filled with concrete
(UHPC) between packers; (c) re-boring providing a stabilized hole with smooth walls [8].
4-6 m
Canister
deposition
holes
Tunnel wall
Clay granule fill compacted on site
Supercontainer in
deployment
mud
Canister
Blocks of highly
compacted
smectite-rich clay
(a)
(b)
(c)
Range is Blue (10 MPa) to Red (60 MPa)
Grey areas > 60 MPa
Highest stresses are near the tunnel oor (72.3 MPa)
Second highest at mid height of the holes (68.3 MPa)
Lowest at bottom of the holes (62.1 MPa)
Range is Blue (10 MPa) to Red (60 MPa)
Grey areas > 60 MPa
Highest stresses are near the tunnel oor (53.2MPa)
Second highest at mid height of the holes (50.1MPa)
Lowest at bottom of the holes (39.3MPa)
deployment
mud

1562

making the placement of super containers containing
the HLW safer, following the principle adopted by
VDH. As for VDH, the containers are preferably
moved down in clay mud for uniform access to water
and early hydration of the clay blocks. This provides
support to the rock and can help to reduce the number
of failed deposition holes when heat is generated.
5.3 VDH (Very Deep Boreholes)
5.3.1 Preparation and Stabilization
One of the necessary steps in preparing the deep
boreholes for waste placement and sealing is to rinse
and clean all parts of the holes and to stabilize the parts
where the holes intersect fracture zones. Where rock
fragments have come off from them the walls are
irregular and the diameter varying, which requires
reaming and casting of concrete followed by re-boring
as indicated in Fig. 15.
5.3.2 Ways of Retarding Expansion of the Clay
Blocks in Super Containers
The very dense clay blocks kept in the containers
when they are moved down in the mud starts expanding
out through the perforation immediately after bringing
the containers into the mud, which begins to
consolidate and ultimately becomes very dense. The
rate of expansion depends on the perforation geometry
and the density of the mud into which the containers
are brought down. The very strong hydrophilic nature
of dense smectite clay blocks make them suck water
from the mud and move out from the containers, which
become surrounded by a successively widening
annulus of dense mud that increases the penetration
resistance at the placement. For minimizing this effect,
the uptake of water from the mud by the dense clay
blocks in the super containers needs to be retarded,
which can be made by coating them with a mixture of
montmorillonite clay and talc (1/3 clay and 2/3 talc).
The dense clay in them firstly sucks water from the thin
coating, which dries and forms thin solid clay plugs in
the perforation holes [26]. When the containers are
moved down in the mud-filled hole, water sucked by
the dense clay has to pass through the plugs, which
delays hydration and expansion of it. After several
hours, the continued expansion of the dense clay causes
disintegration of the plugs and process of water
transport through the thickening mud continues at a
successively slower rate. The properties of the finally
matured clay occupying the space between the
borehole walls and the containers is not affected by the
talc additive because the coating material only makes
up an insignificant part of it.
Another method is to saturate the dense clay blocks
with water before installing them in the super
containers since this reduces the suction potential of the
blocks and strongly retards their expansion and
migration through the perforation [27]. The ultimate
density and properties of the finally matured clay
between the rock walls and the containers will be the
same as when talc-based coatings are used.
5.3.3 Prolonging the Operational Lifetime of Buffer
Clay
While montmorillonite-rich buffer clay appears to
have a long operational lifetime under KBS-3V like
conditions the higher VDH temperatures have led to
questioning of its longevity for serving in the
deployment zones. The experience is that
montmorillonite-rich clay saturated with 10% and 20%
NaCl solutions and then isothermally heated at 110
o
C
under chemically closed conditions for 30 days, retains
its montmorillonite content and physical properties
[28]. However, hydrothermal treatment of such
clayunder open conditions, as in a real VDH, with
exposed to high thermal gradients and a maximum
temperature of 95
o
C gave a reduction of the swelling
pressure by more than 50% and a tenfold increase of
the hydraulic conductivity [29]. The changes were
explained by coagulation of the clay gels and
cementation by silicious precipitates. Saponite and
mixed layer S/I minerals were much less affected by
such hydrothermal treatment making them possible
clay candidate for the manufacturing of dense blocks in
VDH.

1563

(a) (b)
Fig. 16 Copper canisters: (a) type with insert of cast iron containing space for spent fuel (SKB); (b) HIPOW type of 100%
copper with fuel rods (simulated) [7].

Table 3 Assessment of properties and function for the same amount of HLW.
Property/function KBS-3V VDH Note
Constructability Good Good VDH needs study of casing
Rock stability Insufficient Good VDH needs study of mud performance
Installation of waste Difficult Simple KBS-3V needs improvement, VDH needs validation
Isolation of possibly released
radionuclides from the biosphere
Poor Good
Possibly contaminated groundwater moves up in KBS-3V host
rock but stays at depth in VDH host rock
Time for construction > 20 years < 20 years
Parallel construction of several VDHs at each site reduces the
time further
Estimated cost Very high High More detailed calculations are required
Waste retrievability Difficult Possible Very long time required for KBS-3V
Long-term performance Uncertain Good
The impact of seismic and tectonic events is particularly
discouraging for KKBS-3V

5.3.4 Waste Containers
The presently proposed Swedish canister concept for
use in a HLW-repository of KBS-3V type appears in
Fig. 16, which also shows an alternative in the form of
the HIPOW canister that is superior with respect to
the mechanical performance and chemical integrity [7].
It would make disposal of HLW absolutely safe but is
presently not a primary candidate for cost reasons. The
lower construction cost of VDH would allow for
developing the HIPOW technique.
For VDH, which has the stagnant deep-water as
major waste isolation component, simpler and cheaper
canisters and super containers can be considered. Thus,
iron or steel may be acceptable and would also make it
possible to use these metals for preparing the casing.
The oxygen-free water in the deployment zone would
make the corrosion of such canisters very slow and the
larger volume of the corrosion products, mainly
hydroxides, than of the installed canisters and
containers would densify the mud and make it less
permeable. Cementation would, however, make it less
ductile.
6. Conclusions
The KBS-3V and the VDH concepts both require
three well operating barriers for isolating the biosphere
from radionuclides, i.e., suitable host rock, tight
canisters and waste-embedding clay, partly for

1564
different purposes. For the first mentioned concept the
mechanical and hydraulic performances of the rock are
most important while for the latter the rock constitution
and groundwater flow in the upper 2 km are not
essential. The canisters are the most important
waste-isolating barrier for KBS-3V but of less
importance for VDH because of the role of the salt
water to retain possibly released radionuclide at depth.
The waste-embedding clay, finally, is strongly needed
for preserving the KBS-3V canisters by being tight and
ductile, while it plays a minor role for the VDH except
in the upper 2 km. The backfilled deposition tunnels in
a KBS-3V repository provide very limited hindrance of
radionuclides to move to the biosphere while the very
long, dense clay columns in the sealed parts of VDH
effectively prevent Radionuclides to reach up to the
biosphere.
Comparison of the major functions of KBS-3V and
VDH shows that the latter concept has a number of
advantages but that certain issues still remain to be
worked on for accepting it as number one candidate.
The major positive and negative properties are listed in
Table 3.
One concludes from the comparison of the concepts
that KBS-3V, representing various versions of
repositories located at medium depth and VDH,
representing HLW placement at very large depths, that
both are feasible but that VDH has the potential of
being a primary repository concept for crystalline rock.
References
[1] C. Svemar, Cluster Repository Project (CROP), Final
Report of European Commission Contract
FIR1-CT-2000-2003, Brussels, Belgium, 2005.
[2] R.N. Yong, R. Pusch, M. Nakano, Containment of
High-Level Radioactive and Hazardous Solid Wastes
with Clay Barriers, Spon Press, London and New York,
2010.
[3] A Review of the Deep Borehole Disposal Concept for
Radioactive Waste, Nirexreport no. N/108, UK Nirex
Ltd., 2004.
[4] B. Grundfelt, Comparison of the KBS-3 Method and
Deposition in Deep Boreholes for Final Disposal of Spent
Reactor Fuel, Swedish Nuclear Fuel and Waste
Management Co (SKB), R-10-13, Sweden, 2010.
[5] R. Pusch, G. Ramqvist, S. Knutsson, M.H. Mohammed,
The concept of highly radioactive waste (HLW) disposal
in very deep boreholes in a new perspective, Journal of
Earth Sciences and Geotechnical Engineering 2 (3) (2012)
1-24.
[6] R. Pusch, Rock Mechanics on a Geological Base,
Developments in Geotechnical Engineering, Elsevier
Publ. Co., Amsterdam, 1995, p. 158.
[7] R. Pusch, Geological Storage of Radioactive Waste,
Springer-Verlag, Berlin, Heidelberg, 2008.
[8] R. Pusch, G. Ramqvist, N. Bockgrd, L. Ekman, Sealing
of Investigation Boreholes, Phase 4, Final report, SKB
R-11-20 Technical Report, Swedish Nuclear Fuel and
Waste Management Co. (SKB), Sweden, 2011.
[9] R. Pusch, Practical visualization of rock structure, Eng.
Geol. 49 (3-4) (1998) 231-236.
[10] R. Pusch, R. Weston, Superior techniques for disposal of
highly radioactive waste (HLW), Environ Earth Sci. 59
(2012) 75-85.
[11] D. Martin, Brittle rock strength failure: Laboratory and in
situ, in: 8th ISRM Congress, Sep. 25-29, Tokyo, Japan,
1995.
[12] M.D. Zoback, C. Apel, J. Baumgrtner, M. Brudy, R.
Emmermann, B. Engeser, et al., Upper crustal strength
inferred from stress measurements to 6 km depth in the
KTB borehole, Nature 365 (1993) 633-635.
[13] K. Tadokoro, Structure and physical properties of fracture
zone derived from seismic observations at the
Nojima fault and the western Tottori eartquake fault,
Japan. Bull. Earthquake Res. Inst. Univ. Tokyo 78 (2003)
67-74.
[14] S. Hickman, M. Zoback, Stress orientations and
magnitudes in the SAFOD pilot hole, Geophys. Res.
Letters 31 (2004) 1-4.
[15] R. Pusch, O. Karnland, A. Lajudie, A. Decarreau, MX-80
Exposed to High Temperatures and Gamma Radiation,
SKB Technical report TR-93-03, SKB, Sweden, 1993.
[16] H.J. Herbert, J. Kasbohm, H. Sprenger, A.M. Fernndez,
C. Reichelt, Swelling pressures of MX-80 bentonite in
solutions of different ionic strength, Physics and
Chemistry of the Earth 33 (2008) 327-342.
[17] I. Kolakov, J. vandov, R. Pikryl, H. Vinov, V.
Jedinkov-Kov, J. Zeman, Mineralogical changes in
bentonite barrier within Mock-Up-CZ experiment,
Applied Clay Science 47 (2010) 10-15.
[18] R. Pusch, R.N. Yong, M. Nakano, High-Level
Radioactive Waste Disposal, A Global Challenge, WIT
Press, Southampton, Boston, 2012.
[19] R. Pusch, R.N. Yong, M. Nakano, Stiffening of smectite
buffer by hydrothermal effects, Engineering Geology 110
(2010) 21-31.

1565
[20] P. Grindrod, H. Takase, Reactive chemical transport
within engineered barriers, in: Proceeding of 4th Int. Conf.
on the Chemistry and Migration Behaviour of Actinides
and Fission Products in the Geosphere, Charleston, USA,
Dec. 12-17, 1994, pp. 773-779.
[21] D.A. Dixon, J. Graham, M.N. Gray, Hydraulic
conductivity of clays in confined tests under low hydraulic
gradients, Can. Geotechnical Journal 36 (1999) 815-825.
[22] D.M. Anderson, R. Pusch, E. Penner, Physicaland
Thermal Properties of Frozen Ground, in: Geotechnical
Engineering for Cold Regions, McGraw-Hill, UK, 1978,
pp. 44-49.
[23] R. Pusch, L. Warr, G. Grathoff, A. Pourbakhtiar, S.
Knutsson, M.H. Mohammed, A talc-based cement-poor
concrete for sealing boreholes in rock, Engineering 5
(2013) 251-267.
[24] R. Pusch, Waste Disposal in Rock, Developments in
Geotechnical Engineering, Elsevier Publ. Co., 1994, p. 76.
[25] G.G. Meyerhof, Bearing capacity and settlement of
foundations, American Socity of Civil Engneers 102
(1976) 195-228.
[26] R. Pusch, A technique to delay hydration and maturation
of borehole seals of expansive clay, Engineering Geology
121 (2011) 1-6.
[27] T. Forsberg, R. Pusch, S. Knutsson, Rational method for
preparing dense clay blocks with desired degree of water
saturation, Engineering Geology, Technical Note, 2013.
(in print)
[28] J. Kasbohm, R. Pusch, T.L. Nguyen, T.M.T. Hoang,
Labscale performance of selected expandable clays under
HLW repository near-field view, Accepted for Publication
in Environmental Earth Sciences 69 (8) (2012)
2569-2579.
[29] L. Xiaodong, R. Prikryl, R. Pusch, THMC-testing of three
expandable clays of potential use in HLW repositories,
Applied Clay Science 52 (2011) 419-427.


Improvement Properties of Cohesion-Less Soil Using
Recycled Bassanite
Masaki Kobayashi
1
, Aly Ahmed
2
and Keizo Ugai
3

1. Environmental Business Department, DC Co., Ltd., Tokyo 210-005, Japan
2. Civil Engineering Department, Beni-Suef University, Beni-Suef 62512, Egypt
3. Civil and Environmental Engineering Department, Gunma University, Gunma 376-8515, Japan

Abstract: Solid waste management is a serious problem over the world. Therefore, reduction, re-use and recycling of waste have
become major issues in recent days. Gypsum waste plasterboard is considered one example of these waste materials. This study
evaluates the use of recycled bassanite, which is derived from gypsum waste plasterboard, to enhance the performance of two types
of cohesion-less soil. Recycled bassanite was utilized as a stabilizing agent to improve both compressive and splitting strengths of
the tested soil. The effect of bassanite content, soil type, water content and curing time were investigated to explore the behavior of
treated soil with recycled bassanite. Test results showed that increase of bassanite content is associated with increase in optimal
moisture content, while no significant increase in the dry unit weight was observed. Both compressive and splitting tensile strengths
enhanced with the additives of recycled bassanite. The increase of bassanite content had a more significant effect on the compressive
strength compared with the effect on tensile strength. The use of recycled bassanite to enhance the strength of sandy soil had a more
significant effect compared with silty soil. The effect of curing time on the strength of treated samples was more significant in early
curing ages compared with late curing ages. The strength decreased significantly in case of stabilized samples prepared with water
content at the wet-side of the compaction curve. However, insignificant decrease in the strength of the stabilized sample was
detected with moisture content at the dry-side of compaction curve. This research meets the challenges of our society to reduce the
quantities of gypsum wastes, producing useful material from waste materials that will help to a sustainable society.

Key words: Soil stabilization, compressive strength, recycled bassanite, cohesion-less soil, splitting strength.

1. Introduction

Recently, researchers are focused on obtaining a
desirable way to eliminate waste materials by using
them as alternatives in the applications of civil and
industrial engineering. Gypsum waste plasterboard is
considered one example of these waste materials.
Large quantities of gypsum plasterboard,
approximately 80 million tons in the world, are
produced annually. Around 85% of this production
accounts for Europe, USA and J apan. Out of the latter
figure, about 15 million tons is sent to landfill sites [1].
Plasterboard is made from gypsum sheets covered on
both sides with paper sheets. Consequently, increasing
the amount of gypsum waste plasterboard results in

Corresponding author: Masaki Kobayashi, manager,
research fields: geotechnical and environmental engineering.
E-mail: kobayashi_masaki@dccorp.jp.
more quantities, which are disposed in landfills.
Hence, it is deemed necessary to find alternative
solutions to such problems due to the following
reasons: First, gypsum waste plasterboard in landfills
is known under certain circumstances to cause
hydrogen sulfide gasses, which are potentially
harmful for human and animals; Second, there is
currently an increase in the cost of disposal in landfill
and decrease in available ground spaces specified for
landfills sites; Third, landfill regulations and
environmental laws in Japan require that quantities
based on gypsum materials are disposed to landfills,
and that the waste must be sent to the controlled
landfill site, which in turn leads to increase in disposal
cost.
Several researchers have conducted investigations
to improve the physical and mechanical properties of
DAVID PUBLISHING
D
Improvement Properties of Cohesion-Less Soil Using Recycled Bassanite

1567
soil by using different types of waste materials, such
as cement kiln dust, fly ash, bottom ash, blast furnace
slag, stone dust and factory-waste roof shingles [2-6].
Most of the previous studies investigated the use of
recycled gypsum waste plasterboard in agriculture
applications, while a few studies focused on the use of
gypsum wastes in civil and construction engineering
applications. The new unfired gypsum-clay blocks
were manufactured by using processed plasterboard
from recycled sources. The blend of production blocks
comprises 50% of recycled plasterboard. Results
indicated that the product has a good physical
properties compared with natural clay [7]. Laboratory
and field investigations were conducted to develop
effective novel cementation mixes using recycled
gypsum and a range of mineral wastes for road
foundation construction. For this purpose, a series of
laboratory trials were carried out to determine the
optimum proportions of recycled gypsum and mineral
wastes of a binder paste to achieve the highest
compressive strength. The optimum combination for
the novel binder was a mix of 15% recycled gypsum
from waste plasterboard, 5% bypass dust and 80%
basic oxygen slag, which gave a high compressive
strength. Results indicated that plasterboard waste can
be used in low-strength concrete mixes for the
foundations of minor roads and car parks [8].
Generally, recycled bassanite derived from gypsum
waste plasterboard has a potential to be used as a
cementation material because gypsum is the main key
of cementation products. Therefore, the main
objective of this research is to investigate the use of
recycled bassanite derived from gypsum waste
plasterboard as a stabilized agent to improve both
mechanical properties and strength of cohesion-less
soil.
2. Materials and Methods
Three types of materials, which include two
different types of soil, recycled bassanite and portland
cement, were used. The two used soil types were
brought from Gunma prefecture, Japan. The two types
of soil were named as soil-A and soil-B. All physical
and mechanical properties of tested soils are shown in
Table 1. The grain size distribution curves for all the
tested soils are shown in Fig. 1.
Recycled bassanite used herein was derived from
gypsum waste plasterboard. The crushed plaster
gypsum waste (CaSO
4
2H
2
O) was heated at a
temperature ranging from 130
o
C to 160
o
C for a
certain time results in a hemi-hydrate calcium sulphate
(CaSO
4
0.5H
2
O) with three quarters of the water
molecules removed. Firstly, gypsum waste
plasterboard was pulverized and then screened to
remove any impurities such as papers, paints and any
solid material. Afterward, the powder of gypsum
waste was placed on a specified plate and then the
process of heating was started. The preparation of

Table 1 Physical and mechanical properties of tested soils.
Property Soil A Soil B
Effective diameter, (D10), (mm) 0.15 0.06
Nominal mean size, (D50), (mm) 0.21 0.15
Uniformity coefficient, (Cu) 1.53 2.72
Curvature coefficient, (Cc) 0.94 1.02
Passing from sieve No. 200 0.40 11.30
Max. dry unit weight, (kN/m
3
) 15.7 14.28
Optimum moisture content, (%) 13.20 20.40
Specific gravity, Gs 2.60 2.67
Max. void ratio 0.85 1.28
Min. void ratio 0.60 0.79
Soil classification (USCS) SP SP
Soil classification (AASHTO)
A-3
Fine sand
A-2-4
Silty sand

Fig. 1 Grain size distribution curve for different soil types
used.

1568
recycled bassanite was done in some construction
companies in Gunma prefecture, J apan and Fig. 2 is a
schematic diagram for preparing the recycled
bassanite. Four different contents of recycled
bassanite 0, 5%, 10% and 15% based on dry soil
weight were investigated.
Portland cement was used with a content of 3% by
dry soil mass, which was added to test soil sample.
The main reason for the addition of cement is to
prevent the solubility of gypsum when water was
introduced since hydrate calcium sulphate
(CaSO
4
2H
2
O) is a known water-soluble material.
3. Sample Preparation
The effect of using recycled bassanite on the
strength of tested soil was evaluated by conducting a
series of unconfined compression and splitting tensile
strength tests. Cylindrical samples having a 50 mm
diameter and 100 mm height were used. The
cylindrical mold, which was used to form the soil
samples, was made from high grade plastic with 5 mm
thickness to resist any external pressure during the
process of compaction. The reason of using plastic
molds was to prevent the friction between soil
samples and the inner sides of the mold during sample
extraction. In addition, the use of oil to lubricate the
inner sides of the mold was to make sure that no
friction will be occurred during sample extraction.
The samples were placed into the molds and
compacted statically using a static method by
hydraulic Jack to reach the required density. By
knowing mold volume, which is constant for all tests

Fig. 2 Schematic diagram for producing recycled
bassanite from gypsum waste plasterboard [9].
and the desired density, the weight of soil was
determined to validate the required density.
Homogeneity of the samples was of main concern
during preparation and extraction. The soil samples
were extracted from the molds after 24 h. More details
for preparing stabilized soil samples were described
previously [5, 6]. All samples tests were conducted
after seven days of curing at temperature (21 1
o
C)
and relative humidity more than 90%. Some samples
were tested after 3, 14, 28 and 56 days of curing under
the same conditions to investigate the effect of curing
time on the strength of samples treated with recycled
bassanite.
All tests were conducted at the optimal moisture
content and maximum dry unit weight to simulate
field conditions. A standard compaction test was
carried out to determine the dry
density/moisture-content relationships of untreated
and treated soil with different contents of recycled
bassanite, according to ASTM D698-70 specifications.
The oven-dried soil and recycled bassanite mixtures
was firstly dry mixed then the amount of water tap
was added to obtain the moisture content for all tested
soils.
The dry density/moisture-content relations for soils
A and B treated with different contents of recycled
bassanite are shown in Figs. 3 and 4, respectively.
It is clear that the increase of recycled bassanite
content is associated with a significant increase in
optimum moisture content in both the soil types used.
It is attributed to two reasons: Firstly, due to the
tendency of hemi-hydrate calcium sulphate
(CaSO
4
0.5H
2
O) to obtain the three quarters of the
water that was missed before during heating process,
in order to change to hydrate calcium sulphate
(CaSO
4
2H
2
O); The second reason is related to when
fine content such as recycled bassanite was mixed
with soil, more water was required in compaction in
order to achieve the maximum dry unit weight [10]. In

1569

Fig. 3 Dry density/moisture-content curves for tested
soil-A at different contents of recycled bassanite used.

Fig. 4 Dry density/moisture-content curves for tested
soil-B at different contents of recycled bassanite used.

the case of soil type B (silty soil), the increase of
bassanite content up to 15% was associated with a
slightly increase in dry unit weight. This was
attributed to the calcium component in bassanite,
which promotes soil particles to flocculate. Thus,
attraction between soil particles was developed due to
this property of flocculation. It may be assumed that
the flocculation occurred due to the fine components
of the soil which normally carry negative charges and
attract calcium, which carries two positive charges.
This behavior could be related to the tested soil-B
comprising more than 10% content of silty clay,
which can subsequently behave as fine minerals. The
reason of the slightly increase of dry unit weight is
probably related to the increase of optimal moisture
content in the soil-bassanite mixture as compared to
pure soil. Consequently, the maximum dry unit weight
increased slightly. While in the case of soil type A
(fine sand), slight increase in dry unit weight was
related to the particles of bassanite which are very fine
compared to soil particles. That makes it easy for
some bassanite particles to fill the small pores within
the particles of sandy soil. It is difficult from a
practical view to completely fill the pores during
compaction process in the case of pure sand. For the
same volume of compacted sample, the weight of the
bassanite-sand mixture was slightly higher than the
weight of pure sand that was associated with the slight
increase in dry unit weight. On the contrary, if the
difference between soil particles was high such as in
coarse sand, the dry unit weight would be decreased.
The results presented herein are only for soil-B
because the same results were obtained with the other
type of soil-A. The results of unconfined compression
tests, stress-strain relationship, for soil-B type treated
with different contents of recycled bassanite are
shown in Fig. 5. Obviously, the compressive strength
increased with the increase of bassanite content. These
results suggested that recycled bassanite can be used
as a stabilizing material in ground improvement
projects. These results are in agreement with the
previous results [9, 11]. Improvement indexes for both
compressive and splitting tensile strengths against the
content of bassanite are shown in Fig. 5. Improvement
stress index was obtained by dividing the ultimate

Fig. 5 Stress-strain relationships for soil-B with different
bassanite contents.

1570
stress for the treated sample to the ultimate stress of
identical un-treated sample. Fig. 6 indicates that both
compressive and splitting strengths were enhanced
with the increase of recycled bassanite content. The
improvement in compressive strength was more
significant than that in tensile strength. This is
attributed to the fact that stabilized soil cement, which
is weak in tension compared with compression state.
An increase of bassanite content up to 15% enhanced
the improvement of splitting strength compared to
other lower contents used, as shown in Fig. 5. To sum
up, the use of recycled bassanite as a stabilizing agent
to improve the strength of cohesion-less soil can be
acceptable in both compressive and splitting strengths.
Fig. 7 shows the stress-strain relationship for
different two soil types treated with 10% content of
recycled bassanite. It can be observed that the higher
performance was obtained in the case of soil-A
compared with the case of soil-B. The use of bassanite
had a significant effect on the strength enhancement
of sandy soil compared to silty soil. This effect could
be due to the friction between soil particles of sandy
soil was higher than that in silty soil. Besides, the
developed cohesion between particles of sandy soil
was greater than that in silty soil due to the contact
surface area between sand soil particles and bassanite
is strong. Thus, the effect of using bassanite to
enhance the strength of sandy soil was much
pronounced compared to silty sand soil. The same
behavior was obtained in the case of splitting tensile
strength.
To investigate the effect of water content state on
the strength of soil treated with recycled bassanite,
three different moisture contents including OMC
(optimal moisture content) and moisture contents in
dry and wet sides that corresponding to 95% of
maximum dry unit weight, were used, respectively.
The main reason for the use of 95% maximum dry
unit weight was to simulate filed conditions.
Unconfined compression test was conducted on
treated samples of soil-B with recycled bassanite
contents of 5% and 15%, respectively for different
investigated cases of water contents. The suggested
contents of 5% and 15% of recycled bassanite were
representing the smallest and highest content. The
tests were conducted after 7 and 14 days of curing
ages to investigate the effect of water content on the
age of treated samples. The effects of water content on
the strength of treated samples are shown in Figs. 8
and 9. It is clear that for both age of samples, the high
strength was obtained with samples prepared at
optimal moisture content compared to other two used

Fig. 6 Improvement index for both compressive and
splitting tensile strengths in case of soil-B treated with
different bassanite contents.

Fig. 7 Stress-strain relationships for different soil types
used treated with 10% content of recycled bassanite.

Splitting tensile strength
Compressive strength
Bassanite content (%)

1571
water contents. These results are consistent with the
expected routine results of compaction test. The effect
of water content on the strength of 7-day cured
samples was much pronounced than the 14-day cured
samples. The latter difference may be due to the
increase of curing age of the samples. This was
associated with the increase of strength and
consuming most of the excess water as a result of the
reaction between soil particles and bassanite.
Also, it can be noted that the effect of water content
on the samples having highest bassanite content was
more prominent than the samples having smallest
bassanite content. This behavior could be related to
both increase in bassanite content and water content,
as illustrated before. Subsequently, the effect of water

Fig. 8 Ultimate compressive strength against water
content for soil type-B treated with recycled bassanite and
tested after seven days curing age.

Fig. 9 Ultimate compressive strength against water
content for soil type-B treated with recycled bassanite and
tested after 14 days curing age.

content in the case of the samples having the highest
bassanite content was significant especially at
moisture content in the wet-side. Furthermore, these
figures in the case of samples having smallest
bassanite content of 5% showed not much difference
between the strength of samples prepared at moisture
content in dry-side and samples prepared at optimal
moisture content. This result also did not match with
the theory of compaction test because these samples
were impure soil. However, they were treated with
some cementation materials thus their behaviors were
different. In other words, the amount of moisture
content in this case meets the required amount of
water to complete a good reaction between
cementation material and soil particles. Additionally,
the decrease of water content was associated with the
increase of strength. Generally, it was found that the
unconfined compressive strength decreased
significantly in the case of water content at wet-side
that corresponding to 95% of maximum dry unit
weight, while the strength of water content at dry-side
was slightly decreased.
The effect of curing time on the compressive
strength of untreated and treated samples with the
smallest content of 5% and largest content of 15%
was investigated. Consequently, five different curing
times included 3, 7, 14, 28 and 56 days were studied.
As aforementioned, all samples were subjected to
curing under temperature (21 1
o
C) and relative
humidity more than 90%. Fig. 10 shows the effect of
curing time on the compressive strength for tested
soil-B at different curing times. The compressive
strength for all cases, treated and untreated samples,
increased as curing time increased. This finding is
consistent with the theory of curing time for
cementation materials. The proportion of
improvement in the strength of treated samples in the
early curing days was significantly higher than later
curing time days. In other words, the improvement in
strength for treated samples after seven days curing
time was increased slowly in relation with the increase

1572

Fig. 10 Ultimate compressive strength values for soil
type-B tested at different curing times.

of curing time. This behavior could be explained by
the short time required to complete the setting of
soil-bassanite mixture, subsequently the treated
samples regained most of their strength. Hence, the
improvement in strength for treated samples was
evident especially in the first days of curing after
which the rate of improvement was sluggish mainly
with the largest gypsum content. It is worth
mentioning that after 14 days of curing time for both
bassanite contents, the effect of curing time was
insignificant in the context of the strength
improvement. The early strength for soil-bassanite
mixture is actually a vital property in some projects,
such as embankments. This property helps to open
embankments for trafficability after compaction
process directly. Subsequently, the time of
construction decreases and that leads to a reduction in
construction cost.
5. Conclusions
This research meets the challenges of society to
reduce the quantities of gypsum waste plasterboards
and producing useful material from waste materials.
The use of recycled bassanite enhanced the
performance of all the tested soils. Based on test
results, the following conclusions can be drawn:
The compressive and splitting strengths increased
as recycled bassanite content increased. The content of
recycled bassanite had a more significant effect on the
compressive strength compared to the splitting tensile
strength;
The use of recycled bassanite enhanced the
performance of all the tested types of soil. However,
high performance was obtained with only sandy soil;
Unconfined compressive strength decreased
significantly in the case of water content at the wet-side
compared to other cases of optimal moisture and water
content at the dry-side of the compaction curve;
The effect of curing time for samples treated with
recycled bassanite was significant regarding the
improvement of strength, especially in the early curing
days compared with other curing says. The effect of
curing time was not significant after 14 days of curing.
Acknowledgments
Funding of this research was provided through a
grant from J apan Society for the Promotion of Science
(J SPS) under project number P-10377.
References
[1] GRI (Gypsum Recycling International),
http://www.gypsumrecycling.biz/ (accessed May 25,
2013).
[2] Z.A. Baghdadi, M.N. Fatani, N.A. Sabban, Soil
modification by cement kiln dust, Material Civil Eng. J . 7
(4) (1995) 218-222.
[3] G.A. Miller, S. Azad, Influence of soil type on
stabilization with cement kiln dust, Construction and
Building Materials J . 14 (2) (2000) 89-97.
[4] A. Sreekrishnavilasam, S. Rahardja, M. Santagta, Soil
treatment using fresh and landfilled cement kiln
dust, Const. and Building Materials J . 21 (2) (2007)
318-327.
[5] A. Ahmed, M. Shehata, S. Easa, Enhancing the strength
of roadway subgrades using factory-waste roof shingles,
in: The Proceedings of 37th CSCE 2008 Annual
Conference, Qubec, Canada, J une 10-13, 2008, pp.
2546-2555.
[6] A. Ahmed, M. Shehata, S. Easa, Use of factory-waste
shingles and cement kiln dust to enhance the performance
of soil used in road works, Advances in Civil Engineering
J ournal [Online early access]. DOI:10.1155/2009/143750,
Dec. 2009 http://www.hindawi.com/journals/ace/2009
/143750 (accessed Feb. 1, 2013).
[7] WRAP (Waste & Resources Action Programme),
Plasterboard Technical Report: Use of Recycled
Plasterboard in Unfired Clay-Gypsum Blocks, Project

1573
Code: PBD6-403, Banbury, Oxon, UK, 2007,
www.wrap.org.uk (accessed May 25, 2013).
[8] WRAP (Waste & Resources Action Programme),
Plasterboard Case Study: Recycled Gypsum in a Novel
Cementations Material, The Old Academy, 21 Horse Fair,
Banbury, Oxon, UK, 2007, www.wrap.org.uk (accessed
May 25, 2013).
[9] T. Kamei, T. Kato, T. Shuku, Effective use for bassanite
as soil improvement materialsRecycling of waste
plasterboard, Electronic J apanese Geotechnical Journal 2
(3) (2007) 245-252. (In J apanese)
[10] R.D. Holtz, W.D. Kovac, An Introduction to
Geotechnical Engineering, Prentice Hall, New J ersey,
1981.
[11] A. Ahmed, K. Ugai, T. Kamei, Investigation of recycled
gypsum in conjunction with waste plastic trays for ground
improvement, J . Construction and Building Materials 25
(1) (2011) 208-217.


Historical and Future Climatic Change Scenarios for
Temperature and Rainfall for Iraq
Saleh Zakaria
1, 2
, Nadhir Al-Ansari
1
and Seven Knutsson
1

1. Department of Civil, Environmental and Natural Resources Engineering, Lulea University of Technology, Lulea 971 87, Sweden
2. Department of Water Resources Engineering, University of Mosul, Mosul 41002, Iraq

Abstract: The world is facing a big challenge of climatic change, mainly due to increasing concentrations of GHGs (greenhouse gases)
in the atmosphere. Many researches indicated that the climate change occurred disproportionately on developing countries such as
MENA (Middle East and North Africa) countries. The climatic model CGCM3.1 (T47) 2 is used in this research to explain the changes
in average temperatures and the rainfall on the MENA region with special emphases on Iraq. Historical records (1900-2009) and future
(2020-2099) were studied and compared; each period was divided to four sub-periods of thirty years. The results showed that the
average monthly temperature for the four historical periods fluctuated between the lowest and highest value as follows: 9.2-32.9,
10.3-32.7, 9.3-32.8 and 8.6-33.9 (C). The rainfall for historical periods kept on the same distribution during the past 109 years, and
fluctuated between the lowest and highest value of 21.3 mm and 37.6 mm with an average that reached up to 26.51 mm. For the future
period, the maximum average monthly temperature reached up to 37.41 (
o
C) during June and minimum average monthly temperature
reached up to 4.24 (C) during January. The average monthly temperature fluctuated giving a clear impression that the future portends
a higher temperature. The average monthly rainfall, for the future period, fluctuated between the lowest and highest value of 12.91 mm
and 20.63 mm with an average that reached 16.84 mm which represent a reduction percentage of about 36.47% relative to the historical
record of rainfall for the same months.

Key words: Climatic change, MENA, climatic model CGCM3.1(T47) 2, Iraq.

1. Introduction

Climatic change is the biggest challenge facing the
world. The change in the climatic system is mainly due
to the increase in the concentration of GHGs
(greenhouse gases) in the atmosphere. It is believed
that the human activities are responsible for the
increase of GHGs concentrations since the industrial
revolution due to burning of fossil fuels like coal, oil
and natural gas as an energy source.
The impact of climatic change may lead the world to
face a serious risk of dust storms, droughts, floods and
others events which may contribute to food shortages
that inevitably will affect the sustainability of the
communities, and their development.
Climatic change can cause significant impacts on

Corresponding author:

Nadhir Al-Ansari, professor,
research fields: water resources and environmental engineering.
E-mail: nadhir.alansari@ltu.se.
water resources and the hydrological cycle, where the
change of the temperature and precipitation will affect
the evapotranspiration process. Both, quality and
quantity of the runoff component will suffer some
changes which will lead to a significant effect on
sectors like agriculture, industry and urban
development [1].
The prevailing scientific convictions indicated that
the global climate is changing due to human activities
that are affecting three main factors: greenhouse gases,
aerosols and land surface changes [2].
IPCC (intergovernmental panel on climatic change),
in 2001 [3] defined the climatic change as refers to a
statistically significant variation in either the mean
state of the climate or in its variability, persisting for an
extended period (decades). Climatic change may be
due to natural internal processes or external forcing or
to persistent anthropogenic changes in the composition
of the atmosphere or in land use. United Nations
DAVID PUBLISHING
D
Historical and Future Climatic Change Scenarios for Temperature and Rainfall for Iraq

1575
Framework Convention on Climatic Change in 1992
[4], defined the climatic change as a change of climate
which is attributed directly or indirectly to human
activity that alters the composition of the global
atmosphere and which is in addition to natural climatic
variability observed over comparable time periods.
IPCC [5] reported that: the increased in mean annual
global surface temperature, reached 0.74
o
C 0.18
o
C,
most of the warming resulting from human activities
that increase greenhouse gas in the atmosphere.
According to the SRES (Special Report on Emission
Scenarios) for greenhouse gas, the projections of the
temperature at the end of the 21st century range from
1.1
o
C to 6.4
o
C, compared to end-20th century [6, 7].
IPCC [5] observed that there are different impacts on
the physical and biological systems due to climatic
change. The effect of changes in the global average
temperature on global, regional and local levels such
as: ecosystem health, food production, species
distributions and phonology, human health, sea levels,
precipitation and river runoff, drought, average and
extreme changes in temperature and wind patterns. The
impacts are, however, associated with large
uncertainties. Recently, two main methods (adaptation
and mitigation) followed concerning climatic changes.
The first is reducing greenhouse gas emissions, and
the second is the adjustment in natural or human
systems in response to actual or expected climatic
stimuli [5].
Although the international community had identified
the problem of climatic change, it did not succeed to
limit the reasons that are leading to the continuation of
climatic change especially the emission of greenhouse
gases [8]. It is expected that the world will continue to
face increase of the temperature that can reach up to
more than 3 degrees Celsius by the end of this century,
where the concentration of carbon dioxide has risen
from 299.7 ppm (parts per million) in 1910 to 393.9
ppm in March 2012 [9].
The present work based on global climatic
projections and scenarios inspired by national reports
includes a review of the impact of climatic change on
MENA (Middle East and North Africa) region, then it
focuses on Iraq through the distribution of average
monthly temperature and rainfall by historical
(1990-2009) and future (2020-2100) scenarios
according to the climatic change concept using the
results of the Canadian model CGCM3.1(T47),
Scenario A2.
2. Study Area
The MENA region consists of 25 countries covering
an area of 14 million km
2
and 358 million inhabitants
[10, 11]. The MENA land (Fig. 1) extends from the
Atlantic Ocean in the west to the Arabian Sea in the
east, and from the Mediterranean Sea in the north to the
(Africa Horn) and the Indian Ocean in the southeast.
Almost all of the MENA countries are located in

Fig. 1 Map of the MENA countries [14].


1576

semi-arid and arid regions that are highly vulnerable to
climatic change [12]. The annual rainfall at MENA
countries is about 2,282 billion m
3
, while the annual
surface water is about 205 billion m
3
and the annual
groundwater is about 35 billion m
3
[13].
The annual rainfall distribution varies where 52% of
the MENA countries receive an average of less than
100 mm, while 15% receives between 100 mm to 300
mm and 18% receives more than 300 mm. Some
countries like Lebanon, Syria and North African
countries and southern Sudan receive annual rainfall
about 1,500 mm.
MENA countries depend on agriculture and oil for
their economy, therefore they are vulnerable to any
climatic variations, as well as any mitigation policy
aimed at reducing oil.
Iraq is one of the MENA countries located south
west Asian, surrounded by Iran to the east, Turkey to
the north, Syria and Jordan to the west, Saudi Arabia
and Kuwait to the south, and the Gulf to the south-east.
Iraq extends between latitudes 29
o
5' and 37
o
22' north,
and between longitudes 38
o
45' and 48
o
45' east.
The area of Iraq is about 435,052 km
2
. The total
population of Iraq is about 20.4 million (1995), of
which 25% in rural areas. Average population density
is estimated to be 47 inhabitants/km
2
(it is about 5
inhabitants/km
2
, in the desert at the western part
of the country) to more than 170 inhabitants/km
2

in the Babylon province, in the center of the country
[15].
Topographically, Iraq can be divided in four parts:
Mesopotamian plain: Alluvial plain occupies a
quarter of the area of Iraq or equivalent to 132,500
km
2
;
Desert plateau: Located in the west of Iraq and
occupies about less than half the size of Iraq or 168,552
km
2
;
Mountainous region: Mountainous region is
located the northern and the north-eastern part of Iraq
and covers about 92,000 km
2
;
Undulating region: A transition zone between the
low-lying Mesopotamian plain in the south and the
high mountains in the far north and the north-eastern
Iraq and covers 67,000 km
2
.
Iraq lies within the northern temperate zone, but the
climatic is continental and subtropical. Winters are
usually cool to cold, with an average daily temperature
that might reach 16
o
C dropping at night to 2
o
C.
Summers are dry and hot to extremely hot, with a shade
temperature of over 43
o
C during July and August, yet
dropping at night to 26
o
C [11]. The rainfall is tracking
the climate of Mediterranean Sea. Fig. 2 shows
Precipitation Map of Iraq, most of the rainfall falls
during winter, spring and autumn. The rainfall is
non-existent in the summer. Average annual rainfall is
154 mm, and it ranges from less than 100 mm over 60 %
of the country in the south up to 1,200 mm in the
north-east [11, 16]. According to the rainfall
classification of FAO documentation that was
published in 2003 [17], Iraq can be divided into four
agro-ecological zones as follow: (1) Summers prevail
zone located at the northern governorates of Iraq; (2)
Steppes zone has winter rainfall of 200-400 mm
annually. This zone has extremely hot summer and
cold winter, and covers the area located between the
Mediterranean zone and the desert zone; (3) The desert
zone having extreme summer temperatures and less
than 200 mm of rainfall annually, and covers the area
located between north of Baghdad to the Saudi Arabian
and Jordanian borders; (4) The irrigated area, that
extend between the Tigris and Euphrates Rivers, from
north of Baghdad to Basra in the south [17, 19].
Three type of wind below on Iraq. The first is the
northwest wind which blows during all seasons of the
year. In winter, the wind is usually dry cold. While in
the summer, it is smooth and leads to reduce the high
temperatures. The second type of winds is the eastern
or northeastern wind that below in winter and is
accompanied by cold biting and clear sky. Third type
of wind is the southeast wind that is relatively
warm and moist and brings rainy clouds
sometimes.

Fig. 2 Precip

3. Impact
Countries
AFED (
Developmen
countries are
in the world
[20]. The m
average te
precipitation
which alread
and water sc
Haas [21]
change on
conclude tha
evaporation
available wa
Attaher e
changes on
study indica
will increase
Historical a
pitation map o
t of Clima
(Arab Foru
nt) in 2009
e in many wa
to the potent
most signific
emperatures,
n, and SLR
dy suffers fro
carcity.
], projected f
Mediterranea
at the impact
from water
ater supply.
et al. [22] stu
ETO (evapo
ated that proj
e the potentia
and Future Cl
of Iraq [18].
atic Chang
um for En
reported th
ays among the
tial impacts o
cant of whic
less and
(sea level ri
om aridity, r
first order im
an hydrologi
ts may induc
r and soils
udied the im
otranspiration
ected future
al irrigation d
limatic Chang
ge on ME
nvironment
hat the ME
e most vulner
f climatic cha
ch are increa
more err
ise), in a reg
recurrent drou
mpacts of clim
ical systems
ce an increas
that reduce
mpact of clim
n) in Egypt.
climatic chan
demand in Eg
ge Scenarios
ENA
and
ENA
rable
ange
ased
ratic
gion
ught
matic
and
se in
the
matic
The
nges
gypt
by 6
T
dete
reso
hyd
the
situ
ada
[23
T
also
clim
Kuw
o
C)
19 c
T
resu
tem
and
imp
mor
for Tempera
6-16% due to
There is a big
ermine the i
ources and it
drologic regim
MENA reg
uation of loca
aptation meas
].
The Word Ban
o affected by
matic change
wait (52.6
o
C
, Qatar (50.4
countries that
The expected
ult of the imp
mperatures an
d crops; (2)
pacting marin
re intense rai
ature and Rain
o the increase
need for an in
impact of cl
ts componen
mes, drought
gion. This w
al water resou
sures that sho
nk [24], stated
the increase
es. Five ME
C), Iraq (52.0
o
C), and Sud
t set new reco
conditions i
pact of clima
d intense hea
higher air
ne ecosystems
infall, causing
nfall for Iraq

e in ETO by th
ntegrated ana
limatic chang
nts, such as p
t, dam operat
will help to
urces and plan
ould be taken
d that the ME
of temperatu
ENA countrie
0
o
C), Saudi
dan (49.7
o
C)
ord highs in 2
in the MENA
atic change ar
at waves, tha
temperatures
s and fisheries
g both more d
1577
he 2100s.
alysis that can
ge on water
precipitation,
tions, etc. in
identify the
n appropriate
n in advance
ENA region is
ure due to the
es, including
Arabia (52.0
were among
2010.
A region as a
re: (1) higher
at effect lives
s leading to
s; (3) less but
droughts and
7
n
r
,
n
e
e
e
s
e
g
0
g
a
r
s
o
t
d

1578
greater flood
cyclones; (6
and other
increasing ev
negative co
developmen
In recent
climatic in M
by 0.2-0.30
o
precipitation
droughts. F
region will b
the tempera
decade. This
Most of No
will become
Al-Ani [2
elements and
depending o
medium for
results show
impact on
period of the
Jaber [26
proposed cl
model on the
Al-Obeidi
extremism o
in Iraq for t
build a clim
that, three m
moderate an
AL-Bayat
in middle a
(1981-2006)
are located i
aimed to cl
storms and
distribution.
an effect on
most effecti
Historical a
ding; (4) sea
6) new areas
vector and
vidence show
nsequences
nt in the MEN
decades the m
MENA region
o
C per decade
n, strong in s
Future expect
be warmer, d
atures are lik
s is 1.5 times
orth Africa an
e drier [24].
25] analyzed
d variables fo
on the data g
recasts for
wed that the
the temperat
e forecast.
6] studied the
limatic mode
e productivity
i [27] stud
on the environ
the period 19
matic informat
models for rai
nd wet.
ti [28] studie
and southern
) depending
n the western
larify the ge
d the clima
The study sh
n the frequen
ive factor wa
and Future Cl
a level rise;
exposed to
d waterborne
ws that the clim
on the socia
NA countries
most importa
n is the increa
e, rainfall is le
snow mass in
tations are t
drier and mor
kely to rise
faster than th
nd the easter
a number o
or Iraq and su
given by Euro
the period
forecast pro
ture and rai
e climatic ch
els and the
y of wheat an
died the eff
nment of the u
970-2000. Th
tion base. Th
infall varied
d the dust sto
n part of Ira
on five clim
n desert area o
eographic dis
atic factors
howed that e
ncy of this ph
as relative hu
limatic Chang
(5) more inte
dengue, mala
e diseases;
matic change
al and econo
[13].
ant features of
se of tempera
ess, loss of wi
ncluding sum
that the ME
re variable w
0.3-0.40
o
C
he global aver
rn Mediterran
of meteorolog
urrounding ar
opean Centre
1979-1993.
ocess has a c
infall during
hange throug
impact of e
nd barley in Ir
fect of clim
undulating reg
e study aime
he results sho
among semi-
orms phenom
aq for the y
matic stations
of Iraq. The st
stribution of
that affect
leven factors
henomenon.
umidity while
ge Scenarios
ense
aria,
(7)
e has
omic
f the
ature
inter
mmer
ENA
here
per
rage.
nean
gical
reas,
e for
The
clear
the
gh a
each
raq.
matic
gion
ed to
owed
-dry,
mena
years
that
tudy
f the
its
s has
The
e the
eva
effe
3.1
R
a c
imp
seri
and
3.2
IP
of G
gas
and
tota
reac
emi
thou
con
thes
F
Ara
Iraq
T
ME
not
cha

Fig.
wor
for Tempera
aporation an
ective factors
The Impact o
Recent studie
clear impact
pacts of clima
ious risk that
d developmen
Regional Dis
PCC [29] rep
GHGs play an
es such as ca
d others. Duri
al GHGs emi
ched up to 3
issions from a
usand Tg (
ntributed by M
se total world
Fig. 4, show
abia) is contr
q reached up
The amount o
ENA countrie
correspond
ange over the
. 3 Regional d
rld through the
ature and Rain
d clouds q
.
of Climatic C
s indicated th
on climatic
atic change m
affect the life
nt.
stribution of
ported that th
n important r
arbon dioxide
ing the year
ssions from t
25, while the
all resources
(teragram) (
MENA coun
d emissions [3
s that the K
ributing the h
to 100 [30].
of GHGs gase
s represent a
to the proje
region [2].
distribution of
e year 2000 [30
nfall for Iraq
quantity wer
Change
hat greenhous
c change. T
may lead the w
e, health, the
GHGs Emiss
he increase co
role in the clim
e, methane, ni
of 2000, the
the MENA c
e part that co
of the world w
(Fig. 3), th
ntries was ab
30].
KSA (Kingdo
highest perce
es that was em
small proport
ected impacts
f the GHGs em
0].
re the least
se gases have
The negative
world to face a
communities
ions
oncentrations
mate change,
itrous oxide
e total GHGs
countries that
ontributed by
was about 33
he part that
bout 4.2% of
om of Saudi
entage of the
mitted by the
tion and does
s of climatic
missions of the
t
e
e
a
s
s
,
e
s
t
y

t
f
i
e
e
s
c

e

The MEN
high degree
low [31].
WRI [30]
grew 12.7%
a year. How
between 40%
differences i
and develo
produced f
countries w
countries ha
Fig. 5 shows

Fig. 4 MENA

Fig. 5 Globa
Historical a
NA region is
of aridity wh
] reported th
between 200
wever, indiv
% and near z
in scrotal gro
oping countr
from both
with different
ave the greate
s emissions g
A countries co
al CO
2
emissio
and Future Cl
a wide area
here the annu
hat the total g
00 and 2005,
vidual sector
zero, and ther
owth rates be
ries. Emissi
developed
t level. Cert
est share in em
growth by reg
ontributions of
ns growth in se
limatic Chang
characterized
al rainfall is v
global emiss
an average of
rs grew at r
re are substan
etween develo
ions growth
and develop
tainly develo
missions grow
gion in four se
f greenhouse ga
elect sectors: 2
ge Scenarios
d by
very
sions
f 2.4%
rates
ntial
oped
h is
ping
oped
wth.
elect
sect
buil
M
loca
whe
sem
rese
regi
IP
Em
of g
betw
that
as emissions [3
2000-2005 [8].
for Tempera
tors: electric
lding use.
MENA regio
ation that inc
ere most of t
mi-arid and a
earches confi
ions are highl
PCC [5] repo
missions Scena
global GHG
ween 2000 an
t changes in th
30].
ature and Rain
ity and heat

on is influen
creases the im
the region is
arid land zon
irmed the fac
ly vulnerable
orted that the
arios (SRES,
emissions b
nd 2030 (Fig
he global clim
nfall for Iraq
t, transport,
nced by its
mpact of clim
s classified a
nes (Fig. 6) [
ct that arid a
e to climatic c
e IPCC Speci
2000) project
y 25% to 90
g. 7). It will b
matic system w

1579
industry and
geographical
matic change
s hyper-arid,
[32]. Several
and semi arid
change [7].
ial Report on
ts an increase
0% (CO
2
-eq)
be very likely
will be larger
9
d
l
e
,
l
d
n
e
)
y

1580

Fig. 6 Aridit

Fig. 7 Scena
surface tempe

during the 2
continue at o
3.3 The Impa
SLR (sea-
of the clim
century glob
size of seas
Through the
global sea l
during the p
per year [5].
Historical a
ty zones of the
arios for GHG
eratures [3].
21st century a
or above curr
act of Sea-Le
-level rise) is
matic change.
bal warming
and oceans, a
e period 1961-
level rise wa
period 1993-2
.
and Future Cl
world [32].
G emissions fro
and warmer if
rent rates.
evel Rise
one obvious
It is very
has caused a
as well as los
-2003, it was
as about 1.8
2003 it reach
limatic Chang
om 2000 to 21
f GHG emiss
impact as a re
likely that 2
an increase of
ss of land ice
observed tha
8 mm/year w
ed up to 3.1
ge Scenarios
00 (in the abs
sions
esult
20th
f the
[3].
at the
while
mm
M
the
fact
leve
hav
Iraq
area
sea
A
vuln
of I
coa
for Tempera
sence of additi
MENA countr
impacts of S
t that their co
el rise, it will
ve deltaic area
q (Tigris and
as are the mo
level rise [33
AFED [20]
nerable at its
Iraq (Shatt e
astline of Iraq
ature and Rain
ional climatic
ries are part o
SLR with diff
astlines will b
l be more no
as, such as Eg
d Euphrates d
ost vulnerabl
3].
reported th
northern tip n
el-Arab) (Fig
q on the Gu
nfall for Iraq

policies) and p
of vulnerable
fferent levels.
be under the i
oticeable on c
gypt (Nile Riv
delta) (Fig. 8
le in MENA
hat the Gul
north of Kuw
g. 9). Despite
ulf region, th
projections of
e countries to
Despite The
impact of sea
countries that
ver delta) and
). These two
countries by
lf is highly
wait and south
e the limited
he vulnerable
o
e
a
t
d
o
y
y
h
d
e

lowland are
The impact
studied in a
In this asses
used: land, p
urban extent
these impac
ranging from

Fig. 8 The si

Fig. 9 Effect

Historical a
as extend as
t of SLR on
recent assess
ssment, the f
population, G
t, agricultural
cts were ca
m 1 m to 5 m
ituation of the
t of sea level ri
and Future Cl
far inland a
n the MENA
sment by Das
following six
GDP (gross do
l extent and w
alculated for
(Table 1). Da
sea level rise f
ise on Iraq and
limatic Chang
as near Bagh
A countries
sgupta et al. [
x indicators w
omestic produ
wetlands. Fin
SLR scena
asgupta et al.
for Egypt (Nile
d Kuwait [35].
ge Scenarios
hdad.
was
[34].
were
uct),
ally,
arios
[34]
con
sign
and
an i
will
5 m
SLR
pop
e River delta) a
for Tempera
ncluded that
nificant reduc
d 5 m rises, re
impact on 10
l be in the Nil
m SLR. For U
R of 1 m
pulation.
and Iraq (Tigr
ature and Rain
Qatars are
ction of abou
espectively. O
% of Egypts
le Delta, it rea
United Arab
would imp
ris and Euphra
nfall for Iraq
ea would e
ut 2.6 to 13%
One meter SL
s population,
aches a loss o
Emirates an
pact about
ates delta) [33].
1581
experience a
% due to 1 m
LR will cause
more impact
of 20% with a
nd Tunisia, a
5% of the

.

a
m
e
t
a
a
e

1582
Table 1 Impacts of sea level rise: Middle East and North Africa region [34].
Details 1 m 2 m 3 m 4 m 5 m
Area (total = 10,050,556 km
2
)
Impacted area 24,654 33,864 43,727 53,615 63,120
Percent of total area 0.25 0.34 0.44 0.53 0.63
Population (total = 259,396,000)
Impacted population 8,307,472 10,912,744 13,684,993 16,454,655 19,439,678
Percent of total population 3.20 4.21 5.28 6.34 7.49
Urban extent (total = 190,030 km
2
)
Impacted area 3,679 5,037 6,529 7,951 9,384
Agricultural extent (total = 354,294 km
2
)
Impacted area 4,086 6,031 8,007 9,819 11,451
Wetlands area (total = 342,185 km
2
)
Impacted area 11,361 14,758 18,224 21,417 24,277

3.4 The Impact of Freshwater Sources
IPCC [5] stated that the global temperature could be
increased by 3
o
C by the 2050s, if socio-economic
development continues to follow existing trends. One
of the negative impacts of increased temperature is that
it will modify the precipitation patterns leading to
extreme events which will affect the availability of
water resources particularly in tropical and
Mediterranean areas [36, 37]. In most of the MENA
countries the water demand had exceeded the available
water resources which adversely affect the social and
economic development [38].
Most of the MENA water resources are
characterized by low and limited amount. Fig. 10
shows that MENA countries can be divided into three
categories. The first includes Iraq, Sudan and Egypt
which have annual water resources of 75, 65 and 58
billion m
3
/year, respectively. The second includes
Algeria, Lebanon, Mauritania, Morocco, Somalia,
Syria, Tunisia and Yemen with total water resources of
5-30 billion m
3
/year. The remainder has water
resources of less than 5 billion m
3
/year.
Total annual groundwater resources in the MENA
region are about 35 billion m
3
[2, 13].
The combination of the climatic change impact and
rapid trends of population growth, FAO projections
indicated that Algeria, Egypt, Morocco, Syria and
Tunisia are expected to experience severe water
shortages by 2050, and only Iraq is expected to be in a
relatively better situation [31].
It should be mentioned however that Iraq is recently
suffering from reduction in the flow rates of the Tigris
and Euphrates Rivers due to building of huge dams in
Turkey and Syria [11, 16]. However, the water demand
is rapidly increasing for all the MENA countries while
water supply is decreasing.
Climatic change is projected to increase the salinity
levels of ground water as a result of increased
temperature. Furthermore, due to climatic change
conditions the watersheds may face serious soil erosion
and desertification [21].
Recent studies conclude that there are many negative
impacts of climatic change on freshwater systems
represented by increases in temperature, evaporation,
sea level and precipitation variability [7].
The current studies projected that, by the end of the
21st century, the flow of rivers located high latitude is to
be increased while it will be decreasing in the Middle
East, Europe and Central America. However, the
magnitude of these changes is highly uncertain [40].
3.5 The Impact of Increasing Drought
Drought is one of the big problems facing the MENA

Fig. 10 The

region curre
climatic poin
in water or m
normal or ex
hydrological
period of
prolonged fo
hydrological
moisture de
water [41].
Wetherald
both floods
climate with
The drou
water supply
fact, that mo
is rain-fed [4
cereal produ
rain-fed con
It is expe
future as a re
During t
increased in
example in
event every
Historical a
situation of th
ently and in t
nt of view, de
moisture avail
xpected amou
l point of vie
abnormally
or the lack of
l imbalance
eficiency with
d and Manab
and drought
h increased va
ught will aff
y in the MENA
ost of the agri
43]. For exam
uction (wheat
nditions in the
cted that drou
esult of clima
the last 30
n Morocco, T
Morocco, the
5 year perio
and Future Cl
e water resour
the future. Th
efined as tem
lability signif
unt for specif
ew, drought c
y dry weat
f precipitation
, carrying
h respect to
be [42] stated
ts will increa
ariability.
fect the agric
A region [2].
cultural area
mple, about on
and barley) i
e north of Iraq
ught events a
atic change [4
years, dro
Tunisia, Alger
e drought inc
od before 19
limatic Chang
rce for MENA
he drought, f
mporary reduc
ficantly below
fied period. F
can define as
ther sufficie
n to caus a ser
connotation
mans usag
d that the ris
ase due to w
cultural life
This is due to
of MENA reg
ne-third of Ir
s produced un
q [19].
are to increas
44].
ought freque
ria and Syria
creased from
90, to one e
ge Scenarios

countries [39]
from
ction
w the
From
s: a
ently
rious
for
e of
k of
warm
and
o the
gion
raqs
nder
se in
ency
, for
one
vent
eve
eve
and
dec
due
both
3.6
D
cha
in t
prec
lead
A
effe
Iraq
whi
effe
A
emi
indu
hott
amp
on
dus
and

for Tempera
.
ery 2 year peri
ent in the Nea
d Syria were
ades [31]. W
e to the incre
h floods and
The Impact o
Dust storms
ange cloud pr
the polluted r
cipitation wil
ds to produce
AL-Bayati [2
ecting the fre
q, the most im
ile evaporatio
ective.
AFED [20] r
issions in
ustrial emiss
ter and dri
plification eff
Earth will b
t storms in t
d intense.
ature and Rain
iod [45-47]. D
ar East. The r
e the worst
Wetherald and
ease of globa
droughts will
of Dust Storm
may lead to
roperties and
region. In th
ll help to pro
e more dust in
28] indicated
quency of du
mportant fact
on and cloud
reported that
the atmosp
sions, many
ier in a p
ffect, that is, a
become even
the desert wi
nfall for Iraq

Drought is als
recent drough
ever recorde
Manabe [42]
al temperature
l be increased
ms
aerosol poll
then reduce
e desert land
duce more dr
n the air (Fig.
that there ar
ust storms phe
tor was relat
ds quantity w
t due to the
phere from
desert regio
phenomenon
already hot an
more so. C
ill become m
1583
so a recurring
hts in Jordan
ed in recent
], stated that,
e, the risk of
d.
lution which
precipitation
d, the limited
ry soil which
11) [48].
re 11 factors
enomenon in
tive humidity
were the least
e continuous
unregulated
ons will get
called the
nd dry places
Consequently,
more frequent
3
g
n
t
,
f
h
n
d
h
s
n
y
t
s
d
t
e
s
,
t

1584

Fig. 11 A thick dust moving from southwest of Iraq passing Saudi Arabia and then the Red Sea to reach Egypt,
MODIS-Aqua images, May 2005 [33].

3.7 The Impact on Human Life
High temperature adversely affect human life mainly
due to changes in geographical ranges of disease
vectors like mosquitoes, waterborne pathogens, water
quality, air quality and food availability and quality.
Higher CO
2
concentrations and fiercer and more
frequent sand storms in desert areas will increase
allergic reactions and pulmonary diseases all over the
region [5].
Elasha [13] summarized the projected climatic
change impacts on the MENA region depending on the
IPCC assessment and related technical papers, by
dividing the MENA region into three parts: MENA,
Sub-Saharan Countries (Djibouti, Somalia, and Sudan)
and States of the GCC (Gulf Cooperation Council) as
follows:
The mean temperatures of MENA region is
projected to increase 3
o
C to 5
o
C while the
precipitation is to decrease about 20% [5], which lead
to reduction in water runoff by 20% to 30% in most of
MENA region by 2050 [37], and thus the influence
moves to water supply that might be reduced by 10% or
greater by 2050. The weather events will include more
droughts and floods. Mediterranean sea water level is
predicted to rise between 30 cm and 1 m by the end of
the century causing flooding of coastal areas along the
Nile Delta [5].
Warmer temperatures and more variable rainfall are
predicted by climatic models in Sub-Saharan Countries
(Djibouti, Somalia and Sudan) which may lead to more
rapid desertification, droughts, food shortages and
famines. On the other hand, the human health will be
affected by warmer climate, some disease will be more
active like malaria, yellow fever, dengue fever and
other vector-borne. Freshwater deficit problem will
worsen.
Rising both of temperature and sea level will affect
the states of the GCC (Gulf Cooperation Council)
region as follows: increase the salinity of underground
water, more land degradation will occur, biodiversity
on land and in the Gulf will be affected, as well as
coastlines and marine life severely and could impact
desalination plants that are the source of water for the
GCC region [13].

1585
4. Adaptation and Mitigation for Climatic
Change
The IPCC [29] defines adaptation as any
adjustment in natural or human systems in response to
actual or expected climatic stimuli or their effects,
which moderates harm or exploits beneficial
opportunities.
Vulnerability is defined by the IPCC [5], as degree
to which a system is susceptible to, or unable to cope
with, adverse effects of climatic change, including
climatic variability and extremes.
There is a wide range of adaptation that is available
and possible to be followed by the world nations in
order to reduce vulnerability to climatic change.
Part of these adaptation options that can be
implemented in some sectors at low cost, and/or with
high benefit-cost ratios, sectors like water, agriculture,
infrastructure/settlement (including coastal zones),
human health, tourism, transport and energy. There is
high agreement and much evidence of substantial
economic potential for the mitigation of global GHG
emissions over the coming decades that could offset
the projected growth of global emissions or reduce
emissions below current levels [5].
MENA countries will have to play their role
according to the mitigation efforts within international
action, they can adopt development of clean energy
technologies, specifically solar, wind and hydro. As an
example of projects in MENA countries that serve the
climatic change mitigation efforts topic like in Egypt
wind energy, the introduction of CNG (compressed
natural gas) as a transport fuel, the solar power projects,
in Palestine, Tunisia, Morocco and Algeria. In the UAE,
the massive forestation program, zero-carbon city in
Abu Dhabi, in Algeria, the pioneering carbon
capture.
However, the current capacities and actions are
inadequate, most of the above are partially projects that
need support by MENA governments in a unified plan
to be circulated to various MENA countries.
5. Background of Climatic model
CGCM3.1(T47)2
There are several GCMs (global climatic models) as
a research tool, to simulate the climatic change [5].
These GCMs have been developed by different
research groups and organizations over the years. The
output of all has been made available through the IPCC
and its data distribution centre.
SRES (special report on emissions scenarios) is a
report by the IPCC (2000), which contains six models
dealing with climatic changes. The models contain
different scenarios (A1FI, A1B, A1T, A2, B1 and B2).
The SRES scenarios were used in the third and fourth
assessment IPCC reports, they known as (TAR) and
(AR4), where published in 2001 and 2007 respectively.
IPCC did not provide any diagnosis about which
scenario is better or more accurate for the future
emission guess.
The emissions scenario that are used in this work is
(A2) which is a case of rapid and successful economic
development, in which regional average income per
capita converge current distinctions between poor
and rich countries eventually dissolve. The A2
scenarios are of a more divided world than other
scenarios and characterized by: 1/A world of
independently operating, self-reliant nations,
2/continuously increasing population and 3/regionally
oriented economic development.
The dataset was produced by the CRU (CLIMATIC
RESEARCH UNIT) of UEA (University of East
Anglia), and reformatted by IWMI (International
Water Management Institute) [24].
CRU prepared a dataset of mean monthly surface
data (for 1901-2002) over global land areas, for a range
of variables: precipitation and wet-day frequency,
mean, maximum and minimum temperature and
diurnal temperature range, vapor pressure, percent
cloud cover and ground-frost frequency.
The time periods for the historical scenarios in the
CGCM3.1(T47)2 model are across the following
periods: 1900-1930, 1930-1960, 1960-1990 and

1586
1990-2009. While the time periods for the future
scenarios are across the following periods: 2020-2039,
2040-2059, 2060-2079 and 2080-2099.
The main description of the model is as follows:
Model ID: cccma_cgcm3_1, Originating
Group(s): Canadian Centre for Climatic Modeling &
Analysis, Country: Canada, CMIP3 ID: CGCM3.1
(T47)2, Primary Reference: Flato and Boer, 2001;
The model CGCM3.1 (T47) with A2 scenario was
chosen in order to simulate the average monthly
temperature and rainfall on Iraq for the historical
period 1900-2009, and for mean projected temperature
and rainfall till the period 2099.
6. Discussion
The historical period covered 1900-2009, was
divided to four sub-periods as follow: first 1900-1930,
second 1930-1960, third 1960-1990 and fourth
1990-2009. The future period covered (2020-2099),
was also divided into four sub-periods as follow: first
2020-2039, second 2040-2059, third 2060-2079 and
fourth 2080-2099.
6.1 Historical Temperature Change (1900-2009)
Fig. 12 shows the change of average monthly
temperature for the four above periods on Iraq. A
simple comparison for the temperature data indicates
that the average monthly temperature for the four
periods fluctuated between the lowest and highest
value as follows: 9.2-32.9, 10.3-32.7, 9.3-32.8 and
8.6-33.9 (
o
C) within the four periods, respectively.
The first period represents the temperature
benchmark for all periods, the comparison for the first
and second period indicates that the average monthly
temperature for the following six months January,
April, July, September, October and December during
the second sub-period increased. For these months, the
average monthly temperature tend to increase by 0.8,
0.4, 0.4, 0.5, 1.2 and 1.1 (
o
C), respectively. Only two
consecutive months had been increased in temperature
(September and October), while in August and May
there was no increase in temperature. During the
remaining four months, i.e., February, March, June and
November the temperature was decreasing by a range
of 0.4-1.3 (
o
C).
The comparison for the first and third sub-periods
shows that, the third sub-period has increased in
temperature just in five months (January April, August,
October and December) by 1.8, 2.1, 0.2, 1.0 and 0.1
(
o
C), respectively. No consecutive months had been
increased in temperature in third sub-period.
What distinguishes the comparison of the first and

Fig. 12 Change of average monthly temperature for the period 1900-2009 in Iraq.
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
A
v
e
r
a
g
e

m
o
n
t
h
l
y

t
e
m
p
e
r
a
t
u
r
e

(
C
)
Time (month)
1900-1930 1930-1960 1960-1990 1990-2009

1587

fourth sub-periods is that the increase in temperature
has taken place during consecutive months starting in
April and continues until the end of the October with an
increase in temperature which reached up to 1.2, 1.2,
1.0, 0.3, 0.4, 0.7 and 1.4 for these months, respectively.
It seems that the distribution of increase in temperature
differed and took new trend sequentially through these
months.
Fig. 13 shows the average monthly temperature over
the period 1900-2009, where maximum average
monthly temperature reached up to 32.93 (
o
C)
during June and minimum average monthly
temperature reached up to 9.35 (
o
C) during December.
6.2 Historical Rainfall Change (1900-2009)
Fig. 14 shows the change of average monthly rainfall
for the four sub-periods 1990-2009 on Iraq. In general,
the rainfall kept on the same distribution during the
past 109 years, with decline index of amount. A simple
comparison for the rainfall data (Fig. 14) for the four
periods shows that during January-March (first three
months of the year) for the four above-mentioned
sub-periods, the average monthly of rainfall fluctuated
between the lowest and highest value of 21.3 mm and
37.6 mm with an average that reached up to
26.51 mm.

Fig. 13 Average monthly temperature over 109 years (1900-2009).

Fig. 14 The change of average monthly rainfall for the period 1990-2009 in Iraq.
0
5
10
15
20
25
30
35
A
v
e
r
a
g
e

t
e
m
p
e
r
a
t
u
r
e

(
C
)
Time (month)
The average monthly temperature for 1900-2009
The Average monthly temperature for 1900-
0
5
10
15
20
25
30
35
40
45
A
v
e
r
a
g
e

m
o
n
t
h
l
y

r
a
i
n
f
a
l
l

(
m
m
)
Time (month)
1900-1930 1930-1960 1960-1990 1990-2009
The average monthly temperature for 1900-2009

1588

During April, the rainfall starts to decrease where
April represent mid spring season of low rainfall in Iraq.
For this month, the lowest and highest value of the
average monthly rainfall fluctuate between 6.0 mm
and 14.8 mm representing a clear indication of
severe rainfall decrease. For the period May till August,
the average monthly rainfall fluctuates between 0.4-1.2
mm with average that reached up 0.82 mm. During the
last four months of the year (September to December),
the rainfall increased, starting from September, which
is the end of the dry season (summer), by low average
monthly rainfall that was ranging from 5.6 mm to 9.9
mm. The average monthly rainfall begin to increase
starting from October to reach its climax in December,
and on the whole, the average monthly of rainfall
during October to December fluctuate between the
lowest and highest value of (11.3-39.5) mm
respectively with an average that reached 25.66 mm.
Fig. 15 shows average monthly rainfall over 109
year. Rainfall fluctuated between the lowest and
highest values of 0.45 mm (June) and 35.4 mm
(December) respectively. These rainfall values for Iraq
do not satisfy the requirements of successful economic
cultivation of winter crops such as wheat and barley in
the rain-fed farms according to FAO classification.
6.3 Future Temperature Change (2020-2099)
Fig. 16 shows the change of average monthly
temperature for the future sub-periods of 2020-2039,
2040-2059, 2060-2079 and 2080-2099 in Iraq. The
months January, April, July and October represent
mid-seasons for winter, spring, summer and autumn
respectively in Iraq. A comparison for the temperatures
during the mid-seasons during the four sub-periods
shows that the difference in temperature between the
second and first periods reached up to 0.94, 1.86, 1.4
and 0.96 (
o
C) which gives an indication that the reason
of reduction in temperature during movement from
first to second sub-period. The values of difference in
temperature increased during the comparison between
third and first sub-periods that reached up 1.97, 2.93,
2.9 and 2.27 (
o
C). The maximum difference in
temperature appeared during the comparison between
fourth and first sub-periods that reached up to 3.26,
4.28, 4.51 and 4.1 (
o
C).
Giving a clear impression that the future portends a
higher temperature, certainly the reflection of that will
be on the overall facilities of environment.
Fig. 17 shows the future average monthly
temperature over the period 2020-2099, where the
maximum average monthly temperature reached up to
37.41 (
o
C) during June and minimum average
monthly temperature reached up to 4.24 (
o
C) during
January.
6.4 Future Rainfall Change (2020-2099)
Fig. 18 shows the change of average monthly rainfall
for the future sub-periods of 2020-2039, 2040-2059,
2060-2079 and 2080-2099 in Iraq.
During January-March for the four sub-periods, the
average monthly of rainfall fluctuated between the
lowest and highest value of 12.91 mm and 20.63 mm
with an average that reached 16.84 mm which
represent a reduction percentage of about 36.47%
relative to the historical record of rainfall for the same
months. During the month of April, the rainfall
increased in amount, the lowest and highest value of
the average monthly rainfall fluctuates between 18.26
mm and 27.44 mm with average of 22.25 mm while
was 9.85 mm in the historical records.
For the May till August, the average monthly rainfall
fluctuates between 0.05-26.92 mm with average that
reached 6.84 mm, representing a significant increase
compared with its value for the historical record (0.82
mm).
During the last four months of the year
(September-December), the average monthly rainfall
(0.13-0.81 mm) significantly decreased relative to the
historical period of record (5.6-9.6 mm).
The average monthly rainfall begin to increase starting


1589

Fig. 15 Average monthly rainfall over 109 year (1900-2009).

Fig. 16 The change of average monthly temperature for the period 2020-2099 in Iraq.

Fig. 17 Future average monthly temperature for the period 2020-2099.

0
5
10
15
20
25
30
35
40
A
v
e
r
a
g
e

r
a
i
n
f
a
l
l

(
m
m
)
Time (month)
Average monthly rainfall for 1900-2009
0
5
10
15
20
25
30
35
40
45
T
e
m
p
e
r
a
t
u
r
e

(
C
)
Time (month)
2020-2039 2040-2059 2060-2079 2080-2099
0
5
10
15
20
25
30
35
40
T
e
m
p
e
r
a
t
u
r
e

(
C
)

Time (month)
The average temperature 2020-2099
The Average temperatue 2020-2099 The average temperature 2020-2099

1590

Fig. 18 The change of average monthly rainfall for the period 2020-2099 in Iraq.

from October to reach its climax in December
(4.16-23.80 mm) with average that reaches up to 16.13
mm representing significant decrease compared with
the historical period record (25.66 mm).
Fig. 19 shows that the average monthly rainfall over
the years 2020-2099 is fluctuated from 0.55 mm
(September) to 22.59 mm (December), that indicate a
clear further limitation of future rainfall on Iraq.
6.5 The Average Temperature and Rainfall in Iraq
Fig. 20 shows the distribution of average
temperature for each of considered period over the
period 1900-2099.
The annual temperatures reached 21.62, 21.73,
21.58 and 21.61 (
o
C) for the historical sub-periods of
the 1900-1930, 1930-1960, 1960-1990 and 1990-2009,
respectively. For the future sub-periods of 2020-2039,
2040-2059, 2060-2079 and 2080-2099, it reached
18.94, 20.24, 21.39 and 22.85 (
o
C), respectively.
A comparison between these values shows that the
increase of annual temperature was limited during the
historical period 1900-2009, but it was relatively more
significant during the future period 2020-2099. In
general, the above is in agreement with the global
worming that most models considered including the
used model of CGCM3.1 (T47).
Fig. 21 shows that an increase and decrease of
average annual rainfall, over both the historical and

Fig. 19 Average monthly rainfall for the period 2020-2099 in Iraq.
0.0
5.0
10.0
15.0
20.0
25.0
30.0
R
a
i
n
f
a
l
l

(
m
m
)
Time (month)
2020-2039 2040-2059 2060-2079 2080-2099
0.0
5.0
10.0
15.0
20.0
25.0
R
a
i
n
f
a
l
l

(
m
m
)
Time (month)

Fig. 20 Aver

Fig. 21 Aver

future study
the four hist
and 162.6 m
the second s
an increase
reached up 6
sub-periods
with a perce
the third to t
again by 6.
reached 3.73
The avera
from 194.7
reach its low
Historical a
rage annual tem
rage annual ra
y periods. Th
torical sub-pe
mm, respectiv
sub-periods, t
of 12.2 mm
6.68%. Movin
the total rai
entage decrea
the fourth sub
.3 mm, with
3%.
age annual of
mm during
west value of
R
a
i
n
f
a
l
l

(
m
m
)
and Future Cl
mperature ove
ainfall (mm) ov
he average an
eriods was 18
vely. Compar
the total rainf
m. The perc
ng from the se
nfall decreas
se that reache
b-periods the
h a percentag
f the rainfall w
the period o
162.6 mm du
0
50
100
150
200
250
limatic Chang
er the historica
ver the study h
nnual rainfall
82.5, 194.7, 1
ring the first
fall value sho
centage incr
econd to the t
sed by 25.8 m
ed 13.25%. F
total rainfall
ge decrease
was falling d
of 1930-1960
uring 1990-2
Average a
Avera
Avera
Avera
ge Scenarios
al and future st
historical and fu
l for
68.9
and
owed
rease
third
mm,
From
l fell
that
own
0 to
009.
It sh
(ave
mov
T
dec
the
206
T
wer
mm
redu
rain
whe
futu
Time peri
annual rain
age Annual R
Average tem
age temperat
age annual ra
for Tempera
tudy periods.
future periods.
hould be note
erage annual
ving from thi
The right par
rease of the
four sub-p
60-2079 and 2
The average a
re 175.41 mm
m, respective
uced over ti
nfall of the
ere it can rep
ure periods, th
iod (year)
nfall (mm)
Rainfall (mm)
mperature
ture
ainfall (mm)
ature and Rain

ed that the to
l) has toned
ird to fourth s
rt of Fig. 21
average annu
periods of
2080-2099 in
annual rainfal
m, 150.96 mm
ly. The valu
ime. Conside
first future
presents the r
he difference
nfall for Iraq

otal summatio
down its ten
sub-periods.
1 shows an
ual of the rain
2020-2039,
n Iraq.
ll for the four
m, 134.59 mm
ue of total
ering the ave
sub-period
rainfall bench
es in total rain
1591
on of rainfall
ndency when
increase and
nfall through
2040-2059,
r sub-periods,
m and 135.31
rainfall was
erage annual
(2020-2039)
hmark for all
nfall over the
l
n
d
h
,
,

s
l
)
l
e

1592
four sub-periods reached 24.45, 16.37 and then
increased by 0.72 mm, respectively. The percentage of
reduction reached up to 13.93%, 10.83% and then
increased by 0.53%. In general, the average annual
rainfall tends to reduce its values.
7. Conclusions
Most recent studies conclude that wide adaptation is
required to reduce vulnerability to climatic change, a
large array of mitigation options is projected to be
available at least by 2030.
Developing the scientific ability in different fields
related to climatic change should be a top priority,
ensuring political and financial support for the
implementation of adaptation strategies. Increasing
public awareness is very important about the impacts
of climatic change on the environment and human
health. The goal should be a clean technology to reduce
emissions MENA region.
Higher temperatures and less rainfall will make the
MENA countries more arid. These changes will have a
series of effects, particularly on agriculture especially
in rain-fed areas, energy, food security and contribute
to malnutrition, famine and starvation.
Immediate actions should be taken in order to
strengthen the adaptive capacity and reduce the
vulnerability of sensitive sectors and systems to
climatic change.
A wide of strategic policy measures will be required
for adaptation, it is important to focus on
modernization of methods for agricultural sector as
well as enhance food security, expansions in
manufacturing and service industries in rural areas.
The temperature of the world has increased and will
continue to increase over the coming decades coupled
with a decrease in rainfall as shown by CGCM3.1 (T47)
model.
MENA countries must adopt a unified position to
confront the challenge of climatic change and work
with the international community in order to reduce the
impact of climatic change.
Acknowledgments
The authors would like to express their sincere
thanks to Professor Ian Foster of Northampton
University, UK for his suggestions and help during the
preparation of this paper. Thanks to Mrs. Semia Ben
Ali Saadaoui of the UNESCO-Iraq for her
encouragement and support.
The research presented has been financially
supported by Lule University of Technology, Sweden
and by Swedish Hydropower CentreSVC
established by the Swedish Energy Agency, Elforsk
and Svenska Kraftnt together with Lule University of
Technology, The Royal Institute of Technology,
Chalmers University of Technology and Uppsala
University. Their support is highly appreciated.
References
[1] K. Hailemariam, Impact of climate change on the water
resources of Awash River Basin, Ethiopia, climate
research, International and Ultidisciplinary Journal 12
(1999) 91-96.
[2] M. Medany, Impact of Climate Change on Arab Countries,
Chapter 9, 2008, www.afedonline.org/
afedreport/english/book9 (accessed Jan. 1, 2013).
[3] IPCC, Climate change 2001: Synthesis report, in: R.
Watson, the Core Writing Team (Eds.), Cambridge
University Press, Cambridge, UK, 2001.
[4] United Nations Framework Convention on Climate
Change, UNFCCC, Geneva, 1992.
[5] IPCC, Climate Change 2007: The Physical Science Basis,
Cambridge University Press, Cambridge, 2007.
[6] IPCC, Emissions Scenarios Summary for Policymakers,
2000, http://www.ipcc.ch/pdf/special-reports/spm/sres-en.pdf
[7] IPCC, Climate Change 2007: Climate Change Impacts,
Adaptation and Vulnerability, Cambridge University Press,
Geneva, 2007.
[8] Global Carbon-Dioxide Emissions Increase by 1.0 Gt in
2011 to Record, IEA (International Energy Agency), 2012,
http://www.iea.org/newsroomandevents/news/2012 /may/
name,27216,en.html (accessed May 30, 2012).
[9] NOAA, National Oceanic and Atmospheric
Administration, Trends in atmospheric carbon dioxide,
2012, http://www.esrl.noaa.gov/gmd/ccgg/trends/global.html
(accessed June 2, 2012).
[10] World Population Prospects: The 2008 Revision
Population Data Base, UNDESA (United Nations

1593
Department of Economic and Social Affairs), 2009.
[11] N. Al-Ansari, S. Al-Oun, W. Hadad, S. Knutsson, Water
loss in Mafraq Governorate, Jordan, Natural Science 5 (3)
(2013) 333-340.
[12] A.A. Abahussain, S. Anwar, K. Al-Zubari, N. Alaa
El-Deen, M. Abdul-Raheem, Desertification in the Arab
Region: Analysis of current status and trends, Journal of
Arid Environments 51 (2002) 521-545.
[13] B.O. Elasha, Mapping of Climatic Change Threats and
Human development Impacts in the MENA Region,
United Nations Development Programme, MENA Human
Development Report (AHDR), Research Paper Series,
2010.
[14] The Arab Word, http://www.arabbay.com/arabmap.htm
[15] FAO, Water Reports, 1997, pp. 1020-1203,
http://www.fao.org/docrep/W4356E/W4356E00.htm
[16] N. Al-Ansari, S. Knutsson, Toward prudent management
of water resources in Iraq, J. Advanced Science and
Engineering Research 1 (2011) 53-67.
[17] FAO, Arid Zone Forestry: A Guide for Field Technicians,
2003, http://www.fao.org/docrep/ T0122E/t0122e03.htm
[18] T.M. Omer, Country Pasture/Forage Resource Profile,
Iraq, FAO Organization, 2011, http://www.fao.org/ag/
AGP/AGPC/doc/Counprof/Iraq/Iraq.html (accessed Jan. 1,
2013).
[19] FAO, Irrigation in the Middle East Region in Figures,
AQUASTAT Survey 2008, Water report 34, 2008.
[20] Impact of Climate Change on Arab Countries, AFED
(Arab Forum for Environment and Development), 2009.
[21] L. Haas, Mediterranean Water Resources Planning and
Climate Change Adaptation, Mediterranean Regional
Roundtable, Athens, Greece, Dec. 10-11, 2002.
[22] S. Attaher, M.A. Medany, A.A. Abdel Aziz, A. El-Gindy,
Irrigation-water demands under current and future climate
conditions in Egypt, Misr. Journal of Agricultural
Engineering 23 (4) (2006) 1077-1089.
[23] Y. Mohammed, Climate change impact assessment on soil
water availability and crop yield in Anjeni watershed blue
Nilebasin, Master Thesis, Arba Minch, Ethiopia, Aug.
2009.
[24] World Bank, DorteVerner and Mme. Fatma El-Mallah
(League of Arab States), Adaptation to a Changing
Climate in the Arab Countries MNA Flagship report,
Sustainable Development Department, Middle East and
North Africa Region, Oct. 2011.
[25] H. Al-Ani, Tariq, analytical study of some elements and
variables of the climate at Iraq and surrounding areas,
Master Thesis, College of Science, University of
Al-Mustansiriya, 2001.
[26] J.Q. Yahya, Climate change for the semi-mountainous
region in Iraq and its impact on the productivity of wheat
and barley, Master Thesis, Department of Geography,
University of Tikrit, College of Education, 2002.
[27] H. Al-Obeidi, H. Shit, The effect of climate extremism on
the environment of the regions undulating, Doctoral
Dissertation, Department of Geography, College of
Education, University of Tikrit, 2004.
[28] F.F. Al-Bayati, Climatic conditions and their impact on the
geographical distribution of the dust storms Empirical
Study at Al-Anbar province-Iraq, Anbar University,
Journal for the Humanities 1 (2011) 1-26.
[29] IPCC, Climate Change 2001: The Scientific Basis,
Contribution of Working Group I to the IPCC Third
Assessment Report Cambridge, Cambridge University
Press, UK, 2001.
[30] WRI, Earth Trends Data Tables: Climate and Atmosphere,
World Resources Institute, Washington D.C., 2005.
[31] FAO, The State of Food and Agriculture, Food and
Agriculture Organization of the United Nations, Rome,
2002.
[32] Drylands, People, and Ecosystem Goods and Services: A
Web-Based Geospatial Analysis, WRI (World Resources
Institute), 2002.
[33] E. Ghoneim, Remote sensing study of some impacts of
global warming on the Arab Region, in: M. Tolba, N. Saab
(Eds.), Arab Environment Climate Change Report, The
Arab Forum for Environment and Development (AFED),
2009, pp. 31-46.
[34] S. Dasgupta, B. Laplante, C. Meisner, J. Yan, The Impact
of Sea Level Rise on Developing Countries: A
Comparative Study, World Bank Policy Research
Working Paper 4136, Feb. 2007.
[35] Middle East Sea Level Risks, Global Warming Art
License, 2007.
[36] F. Giorgi, Climate Change Hotspots, Geophysical
Research Letters 33, L08707, Implementation of Decision
1/CP.10 of the UNFCCC Convention, 2006.
[37] P.C.D. Milly, K.A. Dunne, A.V. Vecchia, Global patterns
of trends in streamflow and water availability in a
changing climate, Nature 438 (17) (2005) 347-350.
[38] C. Bao, C. Fang, Water resources constraint force on
urbanization in water deficient regions: A case study of the
Hexi Corridor, arid area of NW China, Ecological
Economics 62 (2007) 508-517.
[39] FAO, AQUASTAT database, 2002, http://www.fao.org/
ag/agl/aglw/aquastat (accessed Jan. 1, 2013).
[40] N.W. Arnell, Climate change and global water resources:
SRES scenarios and socio-economic scenarios, Global
Environmental Change 14 (2004) 31-52.
[41] G. Rossi, An integrated approach to drought mitigation in
Mediterranean Regions, in: Proceeding of Drought

1594
Mitigation for the Near East and Mediterranean Workshop,
Syria, 2001.
[42] R.T. Wetherald, S. Manabe, Simulation of hydrologic
changes associated with global warming, J. Geophys. Res.
107 (19) (2002) 43-79.
[43] T. Oweis, A. Hachum, Water harvesting and supplemental
irrigation for improved water productivity of dry farming
systems in West Asia and North Africa, New directions for
a diverse planet, in: Proceedings of the 4th International
Crop Science Congress, Brisbane, Australia, Sep. 26-Oct.
1, 2004.
[44] S. Nicholson, On the question of the recovery of the rains
in the West African Sahel, J. Arid Environ. 63 (2005)
615-641.
[45] M. Karrou, Climatic change and drought mitigation: Case
of Morocco, in: The First Technical Workshop of the
Mediterranean, Component of CLIMAGRI Project on
Climate Change and Agriculture, Rome, Sep. 25-27, 2002.
[46] A. Abbas, Drought suppression procedures for dry lands,
in: The First Technical Workshop of the Mediterranean,
Component of CLIMAGRI Project on Climate Change
and Agriculture, Rome, Sep. 25-27, 2002.
[47] R. Mougou, M. Mohsen, Hendi Zitoune Case Study: Agro
climatic Characterization and Evapotranspiration
Evolution in Climate Change Conditions, Contribution to
the AIACC AF 90 North Africa Project: Assessment,
Impacts, Adaptation, and Vulnerability to Climate Change
on North Africa: Food and Water Resources, Fourth
Tunisian semi-annual report, 2005.
[48] F. Rosenfeld, Y. Rudich, R. Lahav, Desert dust
suppressing precipitation: A possible desertification
feedback loop, Geophysics 98 (11) (2001) 5975-5980.


Correlation between the Bearing Capacity of Crude Oil
Contaminated Soil of Mgbede and the Percentage
Contamination
Iloeje Amechi Francis
Department of Architecture, Enugu State University of Science and Technology, Enugu 400261, Nigeria

Abstract: This study was conducted in Mgbede, River State, Nigeria, hosting up to, or even more than 100 oil wells. It examined the
relationship between the bearing capacity of crude oil contaminated soil and the percentage contamination. Four uncontaminated soil
samples were randomly collected at 1.5 m depth within the oil field with hand auger and analyzed for the load bearing properties
limited to cohesion, angle of internal friction and bulk density. With these parameters, the bearing capacity was determined for each
sample. Crude oil, collected from one of the oil wells with viscosity 0.02611 poises at 40
.o
C and specific gravity 0.8227 g/cm
3
, was
used as the contaminant. This was mixed with the soil sample at 5%, 10%, 15% and 20% concentrations. The mean values of the
bearing capacity were 582.458 KN/m
2
, 495.35 KN/m
2
for square and strip footings respectively at 0% contamination, 240.735
KN/m
2
and 204.753 KN/m
2
at 5%, 321.683 KN/m
2
and 274.593 KN/m
2
at 10%, 127.003 KN/m
2
and 109.12 KN/m
2
at 15%, 105.28
KN/m
2
and 90.758 KN/m
2
at 20% for square and strip footings, respectively. The results showed a consistent decrease in the load
bearing values as the crude oil content increased. The result of the null hypothesis established a strong and significant relationship
between the bearing capacity of crude oil contaminated soil and the percentage contamination.

Key words: Bearing capacity, contaminated soil, crude oil, percentage contamination.

1. Introduction

Since oil exploration and exploitation activities
commenced in Nigeria, dating back to 1958, the host
communities have suffered various forms of
environmental problems involving social, economic,
political, cultural and infrastructural decay in varying
degrees and dimensions. Oil spill is one major source
of these problems as it literally occurs at every stage
of oil production process. When oil spills on land, the
soil is generally contaminated. According to Colorado
Department of Health and Environment [1], OCS (oil
contaminated soil) is any earthen material or artificial
fill that has human or natural alteration of its physical,
chemical, biological or radiological integrity resulting
from the introduction of crude oil, any fraction or
derivative thereof (such as gasoline, diesel or used

Corresponding author: Iloeje Amechi Francis, M.Sc.,
research fields: environmental management/architecture.
E-mail: get2frankfast@gmail.com.
motor oil) or any oil based product. Oil is known to
exert adverse effects on soil properties and plant
community [2].
Spill may be accidental especially during
transportation both on land and sea. It could also be
due to leakage from storage tanks, pipelines or even
during drilling operations. Some cases are attributable
to sabotage or simply purposeful as in the Gulf War in
1991.
When an oil spills or leakage occurs, soils, water
bodies and the air around the sources of leakage are
contaminated. The extent of contamination depends
largely on the chemical composition of the
contaminants and the properties of the soil [3].
In Niger Delta, the intensive oil activities make the
communities highly vulnerable and totally exposed to
the hazards of environmental degradation occasioned
by spilled hydrocarbon. Given the trend of incidents
of oil spillage within the past 55 years of oil
DAVID PUBLISHING
D
Correlation between the Bearing Capacity of Crude Oil Contaminated
Soil of Mgbede and the Percentage Contamination

1596
exploitation in the region, and the steady growth of
more oil sites, oil spillage is expected to be
substantially on the increase in spite of modern
technologies in oil exploration and exploitation
process. Land in these oil rich communities has been
substantially degraded and devastated due to spill.
This degradation resulting from oil and gas production
has attracted the attention of environmentalists and
other experts, Eregha and Irughe [4] quoting from
Ref. [5].
While a lot of studies have been done and
documented on the impact of crude oil spill on soils,
the relationship between the soil strength parameters
and the percentage contamination has not been
sufficiently explored. For instance, in Ref. [6], the
study was limited to shear strength reduction due to
crude oil. In a study carried out by Daniel-Kalia and
Pepples [7] on the Effects of Bony Light Crude Oil
Pollution on soils, the interest was only on plant
growth. Wokocha et al. [8] in their study on the
impact of crude oil spillage on soils, the emphasis was
on food production while Amadi et al. [9]
concentrated on microbiological properties of the soil
in their study on the Chronic Effect of oil spill on soil
properties. According to Refs. [10-12], very few
studies that deal with geotechnical properties of oil
contaminated soil are available in literature. More
studies are therefore very needful, hence this research.
The study will provide a short cut to bearing capacity
evaluation of oil contaminated soil for the study area
using the regression model once the percentage
contamination is determined.
2. Materials and Methods
2.1 Study Location.
The Niger Delta, located within the Atlantic Coast
line of Southern Nigeria, stretching up to 450 km and
terminating at Imo-river entrance, is the worlds
second largest delta [13]. It lies approximately on
latitude 4
o
3
'
to 4
o
50
'
N and longitude 7
o
5
'
to 7
o
35' E,
[14]. Within this region, there are no less than 606 oil
fields with 355 on shore and 251 offshore with 5,284
drilled oil wells and 7,000 km of oil and gas pipelines
[15, 16].
Mgbede in Ogba/Egbema/Ndoni Local Government
Area, which shall constitute the study location, hosts
several oil wells in which diverse oil activities are
carried out on a daily basis. Ogba/Egbema/Ndoni
Local Government Area is made up of three major
ethnic groups, the Ogbas who inhabit the Ogba axis,
the Egbemas at the Egbema axis and the Ndonis
within the Ndoni area, each having its own traditional
ruler [17]. Mgbede is a semi urban oil-bearing
community within the Egbema axis of the local
government area. It has a population of 9,000 people
[18] made up of mainly farmers, artisans and workers
in the companies engaged in oil exploration and
exploitation activities [19]. The study area generally
displays the same kind of geological formation of the
delta areas. It lies within the fresh water alluvial zone
and the Sombrero-Warri Deltaic plain of the Niger
Delta, underlain by the coastal plain sand of the Benin
geologic formation [20]. The soil in this area is
generally a mixture of clay and silt with underlying
coarse sand fractions.
2.2 Experimental Materials and Design
Soil samples were collected from the study area and
analyzed for the load bearing properties. Four
borehole samples of uncontaminated soil were
collected, properly labeled and sent in polythene bags
to the laboratory for the relevant load bearing analyses
which include cohesion, angle of internal friction
(which are shear strength properties) and bulk density.
By simulation, these samples were mixed with
different doses of crude oil (the contaminant) at 5%,
10%, 15% and 20% contamination and tested for the
same properties. The crude oil used was obtained
from the Agip oil well at Mgbede. It is the Sweet
Crude (Bonny Light) with viscosity 0.02611 poises at
40
o
C and specific gravity 0.8227 g/cm
3
.
With the hand auger, holes up to 1.5 m below

1597
ground level were bored and soil samples collected at
that depth. The coordinates of the boreholes were
BH1-604870N, 247133E, 20 m height; BH2-604882N,
247553E, 23 m height; BH3-604917N, 247198E, 23
m height; BH4-604829N, 247203E, 21 m height.
2.3 Hypothesis
H
o
: the bearing capacity of crude oil contaminated
soil does not differ significantly with varying
percentage contamination.
SLR (simple linear regression) technique was used
to test this hypothesis which sought to know whether
the load bearing capacity of the contaminated soil is
significantly related to varying percentage
contamination.
3. Results and Discussions
The following data, presented in Table 1 were
obtained from the analyzed borehole samples.
From the table, the highest bearing capacity values
for both square and strip footings and for all the four
borehole samples were recorded at 0% contamination
while the lowest values were at 20% contamination.
Table 2 presented the summary of the mean values
of the load bearing capacity of the soil samples. A
consistent decrease in the bearing capacity values for

Table 1 Summary of traxial test.
BH N Depth Percentage oil (%) Bulk density
Cohesion
C (KN/m
2
)

o
angle of internal
friction
Bearing capacity
Square footing Strip footing
1
1.5 0 2.20 95 17 579.54 489.95
1.5 5 2.15 85 15 423.77 357.84
1.5 10 2.24 40 7 174.11 148.53
1.5 15 2.23 30 4 111.50 95.42
1.5 20 2.23 25 3 95.42 82.02
2
1.5 0 2.21 54 9 227.60 193.06
1.5 5 2.21 40 4 143.52 122.09
1.5 10 2.21 20 6 97.04 84.26
1.5 15 2.22 17 5 69.62 60.52
1.5 20 2.20 14 3 59.84 52.34
3
1.5 0 2.23 48 17 313.89 268.64
1.5 5 2.14 40 14 214.12 183.12
1.5 10 2.21 37 13 201.06 172.38
1.5 15 2.18 24 9 112.12 96.78
1.5 20 2.09 13 5 55.88 48.92
4
1.5 0 2.16 89 31 1,208.80 1,029.78
1.5 5 2.11 80 28 814.97 693.20
1.5 10 2.19 40 12 214.77 183.76
1.5 15 2.18 39 14 209.98 179.75
1.5 20 2.14 33 11 181.53 155.96
Source: researchers computation.

Table 2 Summary of mean bearing capacity values.
Serial No. Percentage oil (%)
Bearing capacity (KN/m
2
)
Square footing Strip footing
1. 0 582.456 495.350
2. 5 240.735 204.753
3. 10 321.795 274.593
4. 15 127.003 109.12
5. 20 105.280 90.758
Source: researchers computation.

1598

both footing types was noticed as the percentage of
crude oil increased.
This trend shows that increase in crude oil content
reduces the shear strength parameters as the soil
particles will more easily slip or shear with high oil
content. The shear strength of the soil is a crucial
property since it controls the bearing capacity as well
as the stability of the foundation system of a civil
engineering structure [21]. At 0% contamination, the
bearing capacity values for square and strip footings
yielded 582.458 KN/m
2
and 495.350 KN/m
2
,
respectively, and progressively decreased to 105.880
KN/m
2
and 90.758 KN/m
2
at 20% contamination
respectively. Similarly, the values of cohesion (c) and
angle of internal friction (O) were found to be
decreasing with increase in oil content as shown in
Table 1. These are the shear strength parameters.
While angle of internal friction refers to the friction
between the soil particles, cohesion is a measure of
the strength of the bond between these particles. The
decrease in the shear strength, hence, the bearing
capacity, is due to the reduction in the values of the
cohesion and angle of internal friction. This finding is
in line with earlier results obtained by Ref. [22]. The
above relationship was further analyzed statistically
using SLR (simple linear regression) technique to
develop a regression model for the study location.
4. Hypothesis Testing
The hypothesis formulated earlier which sought to
know whether there exists a strong relationship
between the bearing capacity of oil contaminated soil
and varying percentage contamination was tested
using the SPSS (Statistical Package for Social
Sciences) at 95% level of confidence. Data
transformation tool from SPSS was used to obtain the
square root values of all the dependent variables for
the purpose of this regression. ANOVA (analysis of
variance) table was used for further confirmation of
the adequacy of the model.
4.1 Regression Model for Mgbede to Measure
Percentage (%) Contamination against Bearing
Capacity (Square Footing)
In Table 3, the R square value is 0.817, the
ANOVA in Table 4 showed a significant p-value of
0.035.
The regression model for square footing denoted as
MBC
square
becomes:
MBC
square
=22.222 0.64 P (1)
where, MBC
square
= Mgbede bearing capacity for
square footing.
P =Percentage contamination (%).
In Table 5, the standardized beta coefficient is
-0.904 while the significant p-value =0.035
4.2 Regression Model for Mgbede to Measure
Percentage (5) Contamination against Bearing
Capacity (Strip Footing).
In Table 6, for the strip footing the R square value
is 0.814, the ANOVA in Table 7 shows a significant
value of 0.036.

Table 3 Model summary for Mgbede (square footing).
Model R R square Adjusted R square Std. error of the estimate
1 0.904
a
0.817
a
0.756 2.766675
a
Predictor: (constant) percentage (%).
Table 4 Analysis of variance.
Model Sum of squares df Mean square F Significance
Regression 102.356 1 102.3756 13.372 0.035
a

Residual 22.963 3 7.654
Total 125.320 4
a
Predictor: (constant), percentage (%).


1599
Table 5 Coefficients (dependent variable MBC
square
footing).
Model
Un-standardized coefficient
Standardized coefficient
Beta
t Significance
B Std. Error
(Constant) Percentage (%)
22.222
-0.640
2.143
0.175

-0.904
10.369
- 3.657
0.0002
0.035

Table 6 Model summary for Mgbede.
Model R R square Adjusted R square Std. error of the estimate
(Constant) Percentage (%) 0.902
a
0.814 0.752 2.556044
a
Predictors: (constant), percentage (%).

Table 7 Analysis of variance (ANOVA).
Model Sum of squares df Mean square F Significance
Regression 85.978 1 85.978 13.160 0.036
a

Residual 19.600 3 6.533
Total 105.578 4
a
Predictor: (constant), percentage (%)

Table 8 Table of coefficients.
Model
Unstandardized coefficient
Standardized coefficient
Beta
t Significance
B Std. Error
(Constant) Percentage (%)
20.486
-0.586
1.980
0.162

-0.902
10.347
-3.628
0.0002
0.036
a
Dependent variable, MBC
strip
footing.

The regression model for strip footing denoted as
MBC
strip
becomes:
MBC
strip
=20.586 0.586 P (2)
where, MBC
strip
=Mgbede bearing capacity for strip
footing.
P =percentage contamination (%).
In Table 8, the standardized beta coefficient is
-0.902 with a significant p-value of 0.036.
4.3. Decision Rule
The null hypothesis is accepted if the calculated
p-value is greater than 0.05 (p >0.05), otherwise it is
rejected and the alternative, which is the research
hypothesis is accepted.
The result of the only null hypothesis showed that
the calculated p-values in all cases were less than 0.05
(p < 0.05). The research hypothesis is therefore
accepted while the null is rejected. This therefore
suggests that there is a strong and significant
relationship between the load bearing capacity of
crude oil contaminated soil and the degree of
contamination. This is consistent with earlier works in
literature. Evgin and Das [23] in their findings
suggested that settlement of foundation footing would
increase as a result of oil contamination. Also the
findings of Shin and Das [24] indicated that the load
bearing capacity of the soil drops with increase in oil
content.
5. Conclusions and Recommendations
Crude oil pollution reduces the values of cohesion,
angle of internal friction and consequently the bearing
capacity of the contaminated soil. The study also
revealed that as the crude oil content increases there is
a decrease in the strength parameters of the soil
establishing a correlation between the soils bearing
capacity value of crude oil contaminated soil and the
percentage contamination. The researcher therefore
recommends that the design of any structure that
depends on soil for strength and stability in sites prone

1600
to crude oil pollution must consider the reduction
factors generated by the pollutants to ensure that the
soil could still function as a reliable structural support
for the intended load. More studies are however
required to test whether the reduction in bearing
capacity values of oil contaminated soils is
statistically significant.
References
[1] Monitoring and Removal or Treatment of Contaminated
Soil, Colorado Department of Public Health and
Environment, Colorado State USA, 2003.
[2] L.C. Osuji, I. Nwoye, An appraisal of the impact of
petroleum hydrocarbons on soil fertility: The Owaza
experience, African J ournal of Agricultural Research 2 (7)
(2007) 318-324.
[3] P. Fine, E.R. Graber, B. Yaron, Soil interactions with
petroleum hydrocarbons: Abiotic processes, J ournal of
Soil Technology 10 (1997) 133-153.
[4] P.B. Eregha, I.R. Irughe, Oil induced environmental
degradation in the Nigerias Niger Delta: The multiplier
effects, J ournal of sustainable Development in Africa
11(4) (2009) 160-175.
[5] Beyond Scarcity: Power, Poverty and Global Water
Crisis, UNDP report, UNDP (United Nations
Development Programme), 2006.
[6] A.W. Otunyo, Reduction of the shear strength of soils in
the Niger Delta Area of Nigeria due to crude oil production,
Nigerian Journal of Technology 29 (2) (2012) 131-140 .
[7] L.A. Daniel-Kalio, S.F. Pepple, Effects of bonny light
crude oil pollution of soil on the growth of day flower
(Commelia bengha lensis L.) in the Niger Delta, Nigeria,
J ournal of Applied Science and Environmental
Management 10 (2) (2006) 111-114.
[8] G.A. Wokocha, D. Emedu, S. Ihencho, Impact of crude
oil spillage on food production in river state, Nigeria,
J ournal of Money, Investment and Banking19 (2011)
1-10.
[9] A. Amadi, D.A. Samuel, N. Anthony, Chronic effect of
oil spill on soil properties and microflora of a rainforest
ecosystem in Nigeria, Water, Air and Soil Pollution 66
(1994) 59-76.
[10] H.A. Al-Sanad, W.K. Eid, N.F. Ismael, Geotechnical
properties of oil contaminated Kuwait sand, J ournal
Geotechnical Engineering 121 (5) (1995) 407-412.
[11] A.S. Aniban, The effect of temperature on the
engineering properties of oil contaminated sand,
Environment International 24 (1/2) (1997) 153-161.
[12] N.J . Meegoda, P. Ratnaweera, Compressibility of
contaminated fine grained soils, Geotechnical Testing
J ournal 17 (1) (1994)101-112.
[13] L.F. Awosika, Impacts of global climate change and sea
level rise on coastal resources and energy development in
Nigeria, in: J .C. Umulu (Ed.), Global Climate Change,
Impact on Energy Development, DAMATECH Nigeria
Limited, Nigeria, 1995.
[14] H.B. Oku, Spatial variation of soil properties on oil
spillage sites in the Niger Delta, Nigeria, Nigeria J ournal
of Agriculture, Food and Environment 7 (2) (2011)
94-99.
[15] B. Anifowose, Assessing the impact of oil and gas
transport on Nigerias Environment, in: UZI Postgraduate
Research Conference Proceedings, University of
Birmmyham, UK, 2008.
[16] F.C. Onuoha, Oil pipeline sabotage in Nigeria,
dimensions, actors and implication of national security,
L/C, African Security Review Institute for Security
Studies 17 (3) (2008) 99-115.
[17] B.N. Onuebunwa, Garri processing and fish smoking
enterprises: Sustainable development projects in
Ogba/Egbema/Ndoni local government of rivers state,
J ournal of Agriculture and Social Research 4 (2) (2004)
71-77.
[18] Federal Republic of Nigeria, 2006 Population Census,
National Bureau of Statistics, 2006,
http://www.nigerianstat.gov.ng/n bsapps/connections/Pop
2006. pdf (accessed Feb. 15, 2013).
[19] O. Best, O. Chinyere, Experience with the use of
community health extension workers in primary care, in a
private rural health centre institution in south-south
Nigeria, Annals of African Medicine 9 (4) (2010)
240-245.
[20] Environmental Impact Assessment (EIA) of Egbema,
Egbema West and Ugada Integrated Oil and Gas Project,
EIA report, 2010.
[21] H. Rahman, S.N. Abduljauwad, A. Tayyeb, Geotechnical
behaviour of oil-contaminated fine grained soils, response
of Sabkha to laboratory tests: A case study, Journal of
Engineering Geology 33 (2013) 111-125.
[22] B.O. Oluwatobi, M.A. Gbenga, Laboratory studies on the
influence of crude oil spillage on lateritic soil shear
strength: A case study of Niger Delta Area of Nigeria,
J ournal of Earth Sciences and Geotechnical Engineering
3 (2) (2013) 73-83.
[23] E. Evgin, M.B. Das, Mechnical behaviour of oil
contaminated sand, in: Proceeding of the Mediterranian
Conference on Environment Geotechnology, Balkema
Publishers, Rotterdam, The Netherlands, 1992, pp.
101-108.
[24] E.C. Shin, M.B. Das, Bearing capacity of unsaturated
oilcontaminated sand, International J ournal of off
Shore and Polar Engineering 11 (3) (2001) 368-373.

Issue 12, 2013 Journal of Civil Engineering and Architecture

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Issue 12, 2013 Journal of Civil Engineering and Architecture

Загружено:

Авторское право:

Доступные форматы

Journal of Civil Engineering

Вам также может понравиться