Guide to the Design of Concrete Structures in the Arabian Peninsula Appendices
Published by The Concrete Society
Guide to the Design of Concrete Structures in the Arabian Peninsula Appendices Appendices Published by The Concrete Society Published October 2008 © The Concrete Society The Concrete Society Riverside House, 4 Meadows M eadows Business Park, Station Approach, Blackwater, Blackwater, Camberley, Surrey GU17 9AB Tel: +44 Tel: +44 (0)1276 607140 Fax: +44 (0)1276 607141 www.concrete.org.uk Other publications in this series are available from the Concrete Bookshop Bookshop at www.concretebookshop.com Tel: +44 Tel: +44 (0)7004 607777 All rights reserved. Except as permitted under current legislation no part of this work may be photocopied, stored stored in a retrieval system, published, published, performed in public, adapted, broadcast, transmitted, transmitted, recorded or reproduced in any form or by any means, without the prior permission of the copyright owner. Enquiries should be addressed to The Concrete Society. Although The Concrete Concrete Society does its best to ensure that any advice, recommendations recommendations or information it may give either in this publication or elsewhere is accurate, no liability or responsibility of any kind (including liability for negligence) howsoever and from whatsoever cause arising, is accepted in this respect by the Group, its servants or agents. Readers should note that publications are subject to revision from time to time and should therefore ensure that they are in possession of the latest version. Printed by AldenHenDi, Witney, UK.
Guide to the Design of Concrete Structures in the Arabian Peninsula Appendices Appendices Published by The Concrete Society Published October 2008 © The Concrete Society The Concrete Society Riverside House, 4 Meadows M eadows Business Park, Station Approach, Blackwater, Blackwater, Camberley, Surrey GU17 9AB Tel: +44 Tel: +44 (0)1276 607140 Fax: +44 (0)1276 607141 www.concrete.org.uk Other publications in this series are available from the Concrete Bookshop Bookshop at www.concretebookshop.com Tel: +44 Tel: +44 (0)7004 607777 All rights reserved. Except as permitted under current legislation no part of this work may be photocopied, stored stored in a retrieval system, published, published, performed in public, adapted, broadcast, transmitted, transmitted, recorded or reproduced in any form or by any means, without the prior permission of the copyright owner. Enquiries should be addressed to The Concrete Society. Although The Concrete Concrete Society does its best to ensure that any advice, recommendations recommendations or information it may give either in this publication or elsewhere is accurate, no liability or responsibility of any kind (including liability for negligence) howsoever and from whatsoever cause arising, is accepted in this respect by the Group, its servants or agents. Readers should note that publications are subject to revision from time to time and should therefore ensure that they are in possession of the latest version. Printed by AldenHenDi, Witney, UK.
Guide to the Design of Concrete Structures in the Arabian Peninsula Appendices
Contents List of figures List of tables
iv v
Appendix A – Seismic design A1 Introduction A2 Tecto Tectonics nics and seismicity of the region A2.1 The Arabian Plate A2.2 Zagros Belt A2.3 Makran Subduction A3 Regional seismic hazard studies A3.1 Middle East (Mortimer-Lloyd) A3.2 Saudi Arabia (Al-Haddad et al.) al.) A3.3 Iran (Tavakoli and Ghafor-Ashtian Ghafor-Ashtiany) y) A3.4 Europe, Africa and Middle East (Grunthal (Grunthal et al.) al.) A3.5 United Arab Emirates (Abdalla and Al-Homoud) A3.6 Kuwait (Sadek) A3.7 Arabian Gulf Region (Peiris et al.) al.) A3.8 Arabian Gulf (McFarlane) A3.9 United Arab Emirates (Musson et al.) al.) A3.10 Arabian Gulf (Malkawi et al.) al.) A4 Site-specific seismic hazard assessments A4.1 Bur Juman Mall, Dubai (2000) A4.2 Burj Dubai Tower and Mall, Dubai (2003) A4.3 Undisclosed site in Dubai (Sigbjornsson and Elnashai, 2006) A4.4 Discussion of the site-specific seismic seismic hazard hazard assessments A5 Discussion of the seismic hazard for the Arabian Gulf states A5.1 Introduction A5.2 Comparison of PGA estimates A5.3 Comparison of seismic hazard spectra A5.4 Site response effects A5.5 Proposed seismic hazard hazard for for the Arabian Gulf Gulf Region A5.6 Proposed minimum design PGA A6 Conclusions and recommendations A6.1 Summary A6.2 Implications of the inferred UAE West Coast Coast Fault (WCF) A6.3 Seismic hazard hazard for the southern southern Arabian Gulf Region A6.4 Recommended seismic hazard hazard for the Arabian Peninsula
1 1 1 1 1 2 2 3 3 4 6 6 8 8 9 12 12 13 13 14 14 16 17 17 17 18 20 21 23 23 23 23 24
Appendix B – Wind design B1 Overview of wind climate in the Arabian Gulf B2 Characteristics of Arabian Gulf wind systems B2.1 Synoptic winds B2.2 Thunderstorm downbursts B2.3 Arabian Peninsula shamals B2.4 Cyclones (originating in north-western Indian Ocean) B3 Conclusions
25 25 25 25 26 27
Appendix C – Test methods for self-compac self-compacting ting concrete C1 Slump-flow test and T 500 test C2 V-funnel test and V-funnel test at T T 5 5 minutes C3 L-box test method C4 J-ring test
31 31 31 32 32
Appendix D – Cement, slag, pfa and silica fume properti properties es
34
Appendix E – Topography and climate E1 Saudi Arabia E2 Bahrain E3 Kuwait E4 Oman E5 United Arab Emirates E6 The Yemen E7 Qatar
37 37 38 38 39 39 40 40
Appendix F – Permeability testing F1 Permeability tests F1.1 Introduction F1.2 BS EN 12390 Part 8, Depth of penetrati penetration on of water under pressure F1.3 BS 1881 Part 122, Water absorption F1.4 BS 1881 Part 208, Initial surface surface absorption absorption test (ISA (ISAT) T) F1.5 ASTM C1202/AASHTO C1202/AASHTO T-27 T-277, 7, Ability Ability to resist chloride ion penetration (rapid chloride permeability test) F1.6 Nordtest NT Build 443 accelerated chloride penetration F1.7 Nordtest NT Build Build 492 chloride migration from non-steady-state migration experiments F2 Specified values F3 Summary
41 41 41
28 30
42 42 42
42 43 43 43 43
24
iii
List of figures Figure A1 Figure A2 Figure A3 Figure A4 Figure A5 Figure A6 Figure A7 Figure A8 Figure A9 Figure A10 Figure A11 Figure A12 Figure A13 Figure A14 Figure A15 Figure A16 Figure A17 Figure A18 Figure A19 Figure A20 Figure A21 Figure A22 Figure A23 Figure A24 Figure A25 Figure A26 Figure A27 Figure A28 Figure A29 Figure A30 Figure A31 Figure A32 Figure A33 Figure A34
Earthquakes in the Middle East from 1975 to 1995. Seismotectonics Seismotec tonics of the Arabian Plate. Seismic zones and plate boundaries. Comparison of attenuation relationshi relationships. ps. Seismic hazard map for Saudi Arabia showing PGA (g) contours for a 475-year return period. UBC 97 Seismic zonation map for Saudi Arabia. Seismicity of Iran. Seismic source regions for Iran. Acceleration Acceleratio n contours for Iran. GSHAP study for the Middle East. Seismic source regions for the UAE and its environs. UBC 97 Seismic hazard zones and PGS (g) contours for the UAE UAE and its environs. Local seismicity of Kuwait. Seismic hazard map for Kuwait showing PGA contours for a 950-year return period. Seismicity catalogue and seismic source regions. Comparison between UBC 97 and magnitude Mw 7.0 spectra spec tra scaled to short-period (0.4s) accelerations. Comparison between UBC 97 and magnitude Mw 7.0 spectra scaled to PGA. Lower bound 475-year return period PGA and UBC 97 seismic hazard zones for the Arabian Gulf. Upper bound 475-year return period PGA and UBC 97 seismic hazard zones for the Arabian Gulf. Seismic source regions for the UAE and its environs. UBC 97 Seismic hazard zones and PGA (g) contours for the UAE UAE and its environs. Seismic source Zone I. 475-year return period PGA isoseis isoseismals mals for the Arabian Gulf. Seismicity catalogue and seismic source regions for Dubai. Comparison between UBC 97 and Dubai site-specific spectra. Seismicity catalogue. Epicentres (black dots) for for the UAE. Squares show cities. Solid grey lines lines show faults. faults. Solid stars stars mark hypothetical hypothetical epicentres and the white star marks the M 5, 2002, epicentre. Site-specific hazard hazard spectra for Dubai Dubai for return periods of (A) 2475 years and (B) 947 years. years. 2475-year return period seismic hazard spectra for Dubai. Comparison between UBC 97 and Dubai site-specific spectra. Attenuation Attenuati on of of PGA and short-period (0.2s) acceleration acceleration at rock sites for an Mw 5.5 earthquake at reverse fault sources. Attenuation of long-peri long-period od accelerati acceleration on at rock sites for an Mw 7 earthquake ear thquake at reverse fault sources. Comparison of 475-year 475-year return return period bedrock and ground ground surface spectra for for Dubai, 5% damping. damping. Comparison between UBC 97 and 475-year return period site site response spectra for 5% damping. damping.
Figure B1 Figure B2 Figure B3 Figure B4 Figure B5 Figure B6 Figure B7 Figure B8 Figure B9
Radar image of large-scale air mass. Velocity profile of synoptic winds. Typical thunderstor thunderstorm m downburst. Separation of synoptic and thunderstorm data Satellite image of shamal . Shamal velocity velocity profile. Cyclone Gonu (June 2007). Cyclone convective storm system. Cyclone wind speed profiles.
Figure C1 Figure C2 Figure C3 Figure C4
Base plate for flow test. V-funnel. L-box test apparatus. J-ring test apparatus.
iv
List of tables Table A1 Table Table Ta ble A2 Table Ta ble A3 Table Ta ble A4
PGA and spectral spectral accelerations. Comparison of PGA estimates. estimates. Seismic coefficients C a and C v, for UBC 97 Zone 1, the proposed minimum Arabian Gulf Zone, Z AG, and the interpolated 0.06g PGA design spectra. Comparison of PGA estimates estimates for cities in the Arabian Gulf Gulf region.
Table D1 Table Table Ta ble D2 Table Ta ble D3 Table Ta ble D4 Table Ta ble D5 Table Ta ble D6 Table Ta ble D7
Mechanical and physical physical requirements requirements for cement from BS EN 197-1. Chemical requirements requirements for for Portland cement (CEM I) from from BS EN 197-1. Compositional Compositio nal requirements requirements for common common cement types from ASTM ASTM C 150. Physical requirements requirements for for common cement types from from ASTM C 150. BS EN 15167 requirements requirements for for ggbs. BS EN 450 requirements requirements for for fly ash. BS EN 13263 requirements requirements for for silica fume. fume.
Table El Table Table Ta ble E2 Table Ta ble E3 Table Ta ble E4 Table Ta ble E5 Table Ta ble E6 Table Ta ble E7 Table Ta ble E8
Climate in Saudi Saudi Arabia – Jeddah. Climate in Saudi Arabia – Riyadh. Climate in Bahrain. Climate in Kuwait. Climate in Muscat. Muscat. Climate in Sharjah. Climate in Kamaran Island. Climate in Aden, Khormaksar Airport.
Table F1 Table Table Ta ble F2
Ranges of values values for results results of durability tests tests found in various specifications. specifications. Typical values for results results of durability durability tests specified for for adverse exposure conditions. conditions.
v
Appendix A – Seismic design
Appendix A – Seismic design A1
Introduction
This appendix reviews several seismic hazard studies that have been carried out for the Arabian Peninsula. Some of the studies have been deemed to be unreliable and conse-quently have not been included in the recommended level of seismic hazard. Of those studies that are deemed to be reliable, it is recommended that the seismic hazard map presented in Figure A6 be used for most of the Arabian Peninsula, with the exception of those states on the southern shores of the Arabian Gulf. For these states, it is recommended that the seismic hazard map shown in Figure A23 be used, provided that it is supplemented by the minimum design peak ground acceleration (PGA) shown in Table A4.
A2
Tectonics and seismicity of the region
A2.1
The Arabian Plate
Figure A1, taken from the United States Geological Survey (141), shows that the Arabian Peninsula appears to be an area of low seismic risk with most of the seismic activity confined to the boundaries of the Arabian Plate.
The Arabian Plate is bounded by several major fault systems as shown in Figure A2, taken from Johnson (142). The western boundary is the Dead Sea Transform and the eastern boundary is the Owen Fracture Zone. The Arabian Plate is characterised as a zone of tectonic separation from the African Plate along the Gulf of Aden and the Red Sea to the south and south-west. This separation results in a zone of tectonic collision with the Eurasian Plate as the Arabian Plate moves northwards. The most significant earthquakes in the Arabian Gulf are associated with this tectonic collision and they occur in the Zagros Belt and the Makran Subduction, to the north and east of the Arabian Gulf respectively.
A2.2
Zagros Belt
The Zagros Belt is the collision zone between the Arabian and Eurasian Plates. The main Zagros Thrust is located within Iran and it runs parallel to the coastline extending a distance of approximately 1500km, from Turkey in the north to Oman in the south. A zone of severe crustal deformation known as the Zagros Folded Belt extends approximately 200km to the west of the Zagros Thrust. The whole of this region is very active seismically.
N e d u t i t a L
Longitude Figure A1 - Earthquakes in the Middle East from 1975 to 1995.
1
Guide to the Design of Concrete Structures in the Arabian Peninsula
Longitude
N e d u t i t a L
Figure A2 - Seismotectonics of the Arabian Plate.
A2.3
Makran Subduction
The Makran Subduction is where the Gulf of Oman region of the Arabian Plate subducts under the Eurasian plate. This is a region that can produce great earthquakes. However, there is a very distinct difference between the western and eastern parts of the Makran. There is no substantial evidence for large earthquakes in the western portion during the instrumental or historical periods. Furthermore, the eastern Makran, where large earthquakes can happen, is so remote from the Arabian Gulf that it is very unlikely to be the source of appreciable seismic hazard.
2
A3
Regional seismic hazard studies
The seismic hazard studies that have been published for the Gulf States are often contradictory. For example, Al-Haddad et al.(110) suggest that the UAE is almost aseismic, whereas Grunthal et al.(111) suggest that it is an area of high seismic risk. This section describes regional studies that have been carried out in the Gulf States. Most of the historical seismicity catalogues used in these studies are based on work carried out by Ambraseys and Melville(143).
Appendix A – Seismic design
A3.1
Middle East (Mortimer-Lloyd)
This study by Mortimer-Lloyd(144) appears to be based on a previous Building Research Establishment paper by Evans (145). It has been included for historical completeness and it tends to have been superseded by later studies. The study considers the seismic hazard in the Middle East, including Cyprus, Egypt and Libya. It states that a rigorous mathematical treatment of seismic hazard is not possible due to fragmentary evidence. Nevertheless, it does present an indication of the overall seismic risk. The study defines two zones of seismic risk in terms of the Modified Mercalli Intensity (MMI), instead of the more usual peak ground accelerations (PGA). These zones are defined as:
Zone A: Areas in which an MMI intensity of VII or greater could occur within the life-span of a permanent building. The return period is stated to be approximately 60 years. Zone B: Areas which could be affected by larger distant earthquakes or smaller local events giving rise to MMI intensities up to V or VI.
The two zones are based on the following attenuation relationship, taken from Burton (146): MMI = 8.0 + 1.5M – 2.5 ln(h2 + d 2 + 400)0.5 where MMI = M = h = d =
Modified Mercalli Intensity magnitude (Richter) focal depth (km) distance from the epicentre (km).
The resulting seismic risk is shown in Figure A3a. Zones A and B correspond approximately to UBC 97 Zones 2A and Zone 1
45
50
respectively. It is evident that Dubai lies within Zone A, which suggests that the Mortimer-Lloyd(144) study may have been the source reference for the UBC 97 Zone 2A classification by Dubai Municipality. However, the seismic risk in the Gulf States appears to be overestimated for the following reasons: 1.
The boundary of the Arabian Plates is shown passing through the northern UAE, which is approximately 200km westwards of its true position (Figure A3b). The attenuation relationship seems to overestimate intensities at short distances as indicated in Figure A4. In this figure, the Richter magnitude is assumed to be equivalent to the moment magnitude for ease of comparison.
2.
A3.2
This study by Al-Haddad et al.(110) considers the seismic hazard throughout most of the Arabian Peninsula. The seismic source zones defined in this study are located mainly on the western coast along the Red Sea and the Gulf of Aqaba. There are no seismic source zones defined in eastern Saudi Arabia or the other Gulf States, all of which are considered to lack any significant seismic activity. The only source of earthquake activity identified by Al-Haddad et al. that could significantly affect hazard levels in the Gulf States is their source Region 11 in southern Iran, which is assigned a maximum magnitude, Mmax, of 7.5. The 475-year return period accelerations are shown in Figure A5. It is evident that the only significant areas of seismic hazard are the coastal areas along the Red Sea, a small area in north central Saudi Arabia, north-eastern Saudi Arabia, Kuwait and the Iranian coast of the Arabian Gulf.
Incorrect Plate Boundary (Mortimer-Lloyd)
60
55
30
+
Kuwait +
+
+
30
40
45
50
+ +
+ +
+
CYPRUS
+
++
+
+
+ + +
+ + + + ++ + ++
Dhahran BAHRAIN QATAR Hofuf + +
+
Riyadh
+ + ++++ + ++ +
+++ ++ + +
+ +
+
+
35
55
6 0
+
+ +
Unayzah
Saudi Arabia (Al-Haddad et al.)
+
+ + +++ + + + + +
++ +
AD A
UAE
SAUDI ARABIA
B
SYRIA
LEBANON ISRAEL
+
Al Fujayrah
25
Actual Plate Boundary 35 (Johnson)
IRAQ 3 0
JORDAN
Muscat +
BAHRAIN
EGYPT SAUDI ARABIA
20
+
IRAN
R e d
UAE
25
OMAN 2 0
OMAN S e a
(a) Earthquakes and seizmic zones
(b) Discrepancy in plate boundary
Figure A3 - Seismic zones (Mortimer-Lloyd (144) ) and plate boundaries (Johnson (142) ).
3
Guide to the Design of Concrete Structures in the Arabian Peninsula
The choice of attenuation relationship used in this study can be questioned. However, the study has been published in a reputable international journal and the principles used appear to be generally sound. Therefore, the hazard distribution may be considered to be a good indication of the ground motion. Nevertheless, the hazard assessment has been incorporated in the Saudi Arabian building regulations as shown in Figure A6.
A3.3
Iran (Tavakoli and GhaforAshtiany)
This study by Tavakoli and GhaforAshtiany (148) considers the seismic hazard in Iran. The seismicity catalogue used in the study is shown in Figure A7 and the seismic regions used to compute the hazard are shown in Figure A8. The resulting acceleration values (obtained from the study by Grunthal et al.(111) are shown in Figure A9. Figure A4 - Comparison of attenuation relationships (Evans (145) , Mortimer-Lloyd (144) and Boore et al (147) ).
34 32
36
38
40
42
44
46
48
50
52
54
56
58
60 32
5
30 28 26
5 1 . 0
1 . 0 0 .
24 22
30
0 . 2
0 .1
0 0 . 0 . 1 5 0 .1 0 . 2
N e d u t i t a L
0
0 . 1 5
28 0 . 0 5
0 0 . 1 .2 5 0 . 0 0 . 1 5
0 0 . 5
0 .25 0.15 0 .0 5
0 . 0 0 5 . 1
24
0 . 2 0 . 0 5
0 .1 0 0 0 . .1 . 1 2 5
20
0 . 0 5
22
0 . 1 5
0 . 2
0 . 1
20
0 . 0 5 0.15 0.2
18
0 . 1
16
18 0 . 0 5
16
2 . 0
14
34
26
14
36
38
40
42
44
46
48
50
52
54
56
58
60
Longitude Figure A5 - Seismic hazard map for Saudi Arabia showing PGA (g) contours for a 475-year return period (Al-Haddad et al. (110) ).
4
Appendix A – Seismic design
N e d u t i t a L
Longitude Figure A6 - UBC 97 Seismic zonation map for Saudi Arabia (after Al-Haddad et al. (110) ).
40 17 18 20
16
38
19 14
36 N e d u t i t a L
15 9
N e 34 d u t i t a L 32
8
7 13
6
4
11 3
30 12
10
28
5 2
26
24
Longitude Figure A7 - Seismicity of Iran (Tavakoli and Ghafor-Ashtiany (148) ).
1 0
400km
45
50
55
60
Longitude Figure A8 - Seismic source regions for Iran (Tavakoli and Ghafor Ashtiany (148) ).
5
Guide to the Design of Concrete Structures in the Arabian Peninsula
N e d u t i t a L
2
s / m
Longitude Figure A9 - Acceleration contours for Iran (Grunthal et al. (111) ).
A3.4
Europe, Africa and Middle East (Grunthal et al .)
This study was compiled as part of the Global Seismic Hazard Assessment Programme (GSHAP) by Grunthal et al.(111) from individual area studies. The seismic hazard obtained from the GSHAP study, for the Gulf States in general (Figure A9) and for the UAE in particular (Figure A10a), is relatively high. Figures A9 and A10a show hazard levels that reach almost 0.5g in the northernmost part of the UAE and that are greater than 0.2g in most of the country. The implied level for Dubai is of the order of 0.3g. This constitutes a major seismic risk (UBC 97 Zone 3). However, the published UBC 97 zones show the opposite. Doha, Abu Dhabi and Dubai are all specified as Zone 0 and Muscat as Zone 2A. This constitutes negligible and moderate seismic hazard respectively. The probable reason for the high PGA estimate in the GSHAP study is shown in Figure A10b. The Gulf States fell outside the areas covered by the different sub-projects which constituted the study. Grunthal et al. explain that ‘here the hazard was mapped by simulating the attenuated effect of the seismic activity … in the Zagros province of Iran’ . It would appear that the effects of the major earthquakes in Iran were not sufficiently attenuated to obtain reliable estimates of the actual seismic risk in the Gulf States, in general, and in the UAE, in particular. Bommer (149) suggests that these mapped values should be rejected as having a weak scientific basis and being very over-conservative.
6
A3.5
United Arab Emirates (Abdalla and AlHomoud)
This study by Abdalla and Al-Homoud (150) considers the seismic hazard throughout the UAE and its immediate environs. The seismic source regions used in the study are shown in Figure A11 and the resulting acceleration values and UBC 97 zonation are shown in Figure A12. It is possible to criticise any seismic hazard assessment and this study is no exception. The main criticisms are as follows:
The seismic Regions III and VI appear to have been chosen to maximise the risk in the northern UAE, rather than to encapsulate areas of known seismic activity. The maximum earthquake magnitude in the UAE appears to be too high for a region of relatively low seismic activity. For example, Regions I and V, the areas of the highest seismic activity in Iran, were allocated magnitudes of 6.5 and 6.8 respectively. This is lower than the magnitude of 7.0 allocated to Region VII in the UAE. Consequently, the seismic risk in the UAE appears to be overestimated by extending the Zagros hazard levels southwards into the UAE.
The latter of these criticisms was addressed in a site-specific study for Dubai (see Al-Homoud(151)), in which the magnitude for Region VII was reduced to 5.5. This resulted in a reduction in the seismic hazard for Dubai from UBC 97 Zone 2A to below Zone 1. This estimate of seismic hazard correlates with the independent studies carried out by Bommer (149) , Wood and Irvine (152), Peiris et al.(116), McFarlane(153) and Musson et al.(117).
Appendix A – Seismic design
6 15 10 11
40
14 30 0
0
12
0
20
N e d u t i t a L
16
13
10
0
(a) Seismic hazard
(b) Areas covered by local studies
Figure A10 - GSHAP study for the Middle East (Grunthal et al. (110) ).
N e d u t i t a L
Longitude
Figure A11 - Seismic source regions for the UAE and its environs (Abdalla and Al-Homoud (150) ).
7
Guide to the Design of Concrete Structures in the Arabian Peninsula
N e d u t i t a L
Longitude
Figure A12 - UBC 97 S eismic hazard zones and PGS (g) contours for the UAE and its environs (Abdalla and Al Homoud (150) ).
A3.6
Kuwait (Sadek)
This study by Sadek (154) considers the seismic hazard for Kuwait. The local seismicity used in the study is shown in Figure A13 and the resulting acceleration values are shown in Figure A14. It should be noted that the acceleration values used are based on a 10% probability of exceedence in 100 years. This equates to a 950-year return period whereas the other studies in this paper use a 475-year return period.
A3.7
Arabian Gulf Region (Peiris et al.)
A probabilistic seismic hazard assessment for the Arabian Gulf Region was carried out by Peiris et al.(116). Their seismicity catalogue and source zones are shown in Figure A15. The seismicity catalogue included earthquakes since 0 AD, with foreshocks and aftershocks removed.
PGA (g)
0.225g 0.200g 0.175g 0.150g 0.125g 0.100g 0.075g 0.050g
Figure A13 - Local seismicity of Kuwait (Sadek (154) ).
8
Figure A14 - Seismic hazard map for Kuwait showing PGA contours for a 950-year return period (Sadek (154) ).
Appendix A – Seismic design
Figure A15 - Seismicity catalogue and seismic source regions (Peiris et al. (116) ).
Peiris et al. state that the seismic hazard for most cities in the Arabian Gulf Region is similar. The sole exception is shown to be Kuwait City which has a higher seismic hazard. Peiris et al. derived PGA and short-period (0.2s) and 1s period spectral accelerations for several cities in the region as show in Table A1. Location
PGA (g)
0.2s S A (g)
1.0s SA (g)
475-year 2475-year 475-year 2475-year 475-year 2475-year Abu Dhabi 0.05
0.10
0.081
0.145
0.038
0.074
Dubai
0.06
0.12
0.107
0.186
0.053
0.102
Kuwait City 0.12
0.27
0.125
0.298
0.044
0.082
Manama
0.06
0.12
0.089
0.167
0.039
0.071
Doha
0.05
0.10
0.081
0.147
0.035
0.065
Muscat
0.04
0.10
0.092
0.186
0.048
0.121
Table A1 - PGA and spectral accelerations (after Peiris et al. (116) ).
The seismic hazard derived by Peiris et al. is discussed later.
A3.8
Arabian Gulf (McFarlane)
A deterministic seismic hazard assessment was carried out by McFarlane(153) for the southern Arabian Gulf States. This procedure is approximate, but it is simpler and more transparent than a full probabilistic analysis. The procedure may also be used as a check of the results obtained from probabilistic analyses. It is similar to the methodology suggested by Bommer et al.(155) and Bommer(149). This method defines an earthquake scenario (magnitude and distance) that is comparable with the seismic hazard at a site from a seismic source region.
The main seismic hazard for most of the Gulf States is from earthquakes in the Zagros Belt in Iran. Therefore, the 475-return period seismic hazard is adequately represented by an Mw 7.0 earthquake in source region 13 (Zagros Belt) defined by Tavakoli and Ghafor-Ashtiany (148). McFarlane modelled the seismic hazard in the Gulf States by calculating the acceleration at various distances from region 13 to determine the appropriate isoseismals. The attenuation relationship developed by Boore et al.(147) was used. The response spectra derived using this attenuation relationship are shown in Figures A16 and A17 and they are compared with the UBC 97 Zone 2A and Zone 1 spectra. McFarlane originally calibrated the isoseismals to short-period (0.4s) accelerations, which gave a good approximation to the UBC 97 spectra as shown in Figure A16. However, it is more conservative to calibrate the seismic hazard to PGA, which results in the spectra shown in Figure A17. The resulting seismic hazard is shown in Figure A19. The isoseismals were derived by placing Mw 7.0 events along the border of the Tavakoli and Ghafor-Ashtiany (148) source region 13. The PGA isoseismals for 0.15g, 0.75g and 0.05g are located at 42km, 102km and 170km from source region 13. These correspond to UBC 97 Zone 2B, Zone 2A and Zone 1 respectively. The seismic hazard described in Figures A18 and A19 may be regarded as reasonable upper and lower bounds of the actual seismic hazard.
9
Guide to the Design of Concrete Structures in the Arabian Peninsula
Figure A16 - Comparison between UBC 97 and magnitude M w 7.0 spectra scaled to short-period (0.4s) accelerations.
Figure A17 - Comparison between UBC 97 and magnitude M w 7.0 spectra scaled to PGA.
10
Appendix A – Seismic design
Figure A18 - Lower bound 475-year return period PGA and UBC 97 seismic hazard zones for the Arabian Gulf (McFarlane (153) ).
Figure A19 - Upper bound 475-year return period PGA and UBC 97 seismic hazard zones for the Arabian Gulf.
11
Guide to the Design of Concrete Structures in the Arabian Peninsula
30
50
51
52
53
54
55
56
57
58
30 0 . 2g
7.0
6.4
29
Iran
g 3 . 0
29
6.4
Khasab Zone 2A 6.6
28
28
7.0
7.1
0. 1 g
0 . 0 5 g
Ras Al Khaimah
5.7
27
Doha
27
6.4
6.5
Umm Al Qiwen Ajman Zone 1 Sharjah 0 . 0 Dubai 5
ARABIAN GULF
Quatar
5.4 5.8
4.9
5.8
5.8
26
Oman Dibba Zone 2A
0. 1 g
g
Fujairah GULF OF OMAN
26
5.1
Abu Dhabi
4.8
Al Ain
25
25
24 N
3.8
24
24
Zone 0
5.5
23
26 N
Oman
23
5.5
Saudi Arabia 22
50
51
52
53
54
55
56
57
58
22
Figure A20 - Seismic source regions for the UAE and its environs (Musson et al. (117) ).
A3.9
52 E
54 E
56 E
Figure A21 - UBC 97 S eismic hazard zones and PGA (g) contours for the UAE and its environs (Musson et al. (117) ).
United Arab Emirates (Musson et al.)
A probabilistic seismic analysis for the UAE and surrounding area was carried out by Musson et al.(117). The seismic source regions used in the study are shown in Figure A20 and the resulting acceleration values and UBC 97 zonation are shown in Figure A21. The seismic hazard derived by Musson et al. is discussed later.
A3.10
Arabian Gulf (Malkawi et al.)
A probabilistic analysis for the UAE and its immediate environs was carried out by Malkawi et al.(112). They did not split the region into specific seismic source zones based on k nown tectonic features. Instead they used two mega-zones as seismic source regions. The theoretical exactitude of using mega-zones for a probabilistic analysis could be q uestioned; however, the results from the assessment are very similar to those studies that use theoretically correct source-region methodology. The megazones used by Malkawi et al. are Source I (which is shown in Figure A22) and Source II which included Source I and extended it to an area approximately 1000km from the UAE. They state that the two source zones give approximately the same results, but that Source I gives slightly higher PGA. They also state that this may be due to the smaller activity rate parameter for Source I.
Figure A22 - Seismic source Zone I (Malkawi et al. (112) ).
However, Source II gives results that are closer to the deterministic assessment carried out by McFarlane(153) and the probabilistic assessments by Peiris et al.(116) and Musson et al.(117). Furthermore, the Source II assessment gives an excellent correlation with the site-specific assessments carried out by Wood and Irvine (152) and Al-Homoud(156) for Dubai, which are described in the next section. Therefore, the Source II assessment will be used in this report. The seismic hazard derived by Malkawi et al. from this source is shown in Figure A23.
12
Appendix A – Seismic design
Figure A23 - 475-year return period PGA isoseismals for the Arabian Gulf (Malkawi et al. (112) ).
A4
Site-specific seismic hazard assessments
Four site-specific hazard assessments for Dubai, UAE are presented in this section. They were carried out independently by seismologists who used different analysis methodologies and source-zone modelling.
A4.1
Bur Juman Mall, Dubai (2000)
In Dubai, seismic design to UBC 97 Zone 2A is required for buildings over four storeys. However, the proposed extension to the existing Bur Juman Centre would increase the height of the existing four-storey building to five storeys. This would entail an upgrade of the building to comply with seismic detailing to UBC 97 Zone 2A. Therefore, a seismic assessment was carried out by Bommer(149) in order to obtain a relaxation in the building regulations by reducing the seismic detailing for the upgrade works to UBC 97 Zone 1. This would make the upgrade work less onerous. In the seismic hazard assessment, Bommer states that the region surrounding Dubai is almost aseismic. He also states that the main seismic hazard is due either to small local earthquakes that are of little structural significance or to large seismic events in the Zagros region on the Iranian side of the Arabian Gulf.
A deterministic approach was used in this assessment and maximum credible earthquakes at known seismic zones were derived from the Tavakoli and Ghafor-Ashtiany (148) study. The earthquake scenarios used were magnitude Mw = 7.0 at 95km from Dubai and magnitude Mw = 7.2 at 160km from Dubai in the Zagros Folded Belt. Bommer shows that an M = 7.0 in Region 13 of the Ghafor-Ashtiany study is the critical design case for Dubai. A typical design spectrum, derived by Bommer, is shown in Figure A25. The following conservative assumptions were made in the hazard assessment: 1.
2.
3.
The maximum credible earthquake in the seismic source zone defined for the region was used to establish the earthquake scenario. The spectral ordinates were calculated using an attenuation relationship (Boore et al.(147)) that accounts for the increased amplitude in ground motions due to reverse faulting earthquakes, such as those encountered in the Zagros Belt. The site-specific spectra were calculated using the meanplus-one-standard-deviation values from the attenuation relationship, which corresponds to the 84-percentile.
Bommer concluded that the Zone 2A spectrum is unnecessarily conservative for the level of seismic hazard at the site. He also recommended that the UBC 97 Zone 1 spectrum could be safely and confidently adopted for seismic design in Dubai.
13
Guide to the Design of Concrete Structures in the Arabian Peninsula
N e d u t i t a L
Longitude
Figure A24 - Seismicity catalogue and seismic source regions for Dubai ( Wood and Irvine (152) ).
A4.2
Burj Dubai Tower and Mall, Dubai (2003)
Two studies were used for the seismic design of the Burj Dubai project, namely Wood and Irvine (152) and Al-Homoud (151). Both studies used a probabilistic approach to assess the seismic hazard. The seismicity catalogue and the source regions used by Wood and Irvine are shown in Figure A24. The resulting hazard is discussed later. An earlier study by Al-Homoud(151) used a similar seismicity catalogue and identical source regions to the Abdalla and AlHomoud study (150) . These studies contain the error discussed previously, namely the use of an unrealistically large earthquake magnitude in the Northern Emirates. Consequently, the study was rejected by the consulting engineer as unreliable and providing an overestimate of the seismic hazard. A revised study was carried out by Al-Homoud (156) for the Dubai Mall project. In this study the maximum earthquake magnitude was reduced from 7.0 to 5.5 in source region VII shown in Figure A11.
14
This resulted in a reduction of the seismic risk from UBC 97 Zone 2A to Zone 1. Figure A25 shows that the resulting spectrum compares very well with the Wood and Irvine (152) spectrum. It also compares well with the deterministic spectrum suggested by a Bommer (149).
A4.3
Undisclosed site in Dubai (Sigbjornsson and Elnashai, 2006)
Sigbjornsson and Elnashai (157) carried out this assessment for an undisclosed site in Dubai during April 2004. The s eismicity catalogue used by them is shown in Figure A26. They indicate that they have used the seismic regions described in the Tavakoli and Ghafor-Ashtiany (148) assessment, but they also indicate that they have also included the Dibba Fault on the east coast of the UAE and a fault along the west coast of the UAE. However, no details of the seismic source regions or hazard calculations are presented in the paper.
Appendix A – Seismic design
Figure A25 - Comparison between UBC 97 and Dubai site-specific spectra.
The UAE West Coast Fault (WCF) is shown in the tectonic map of Saudi Arabia (Johnson (142)) and the source for the fault trace on the map appears to be Murris (158). However, there is no direct reference to the WCF in either Johnson or Murris, but the latter does interpret it to be part of the Dibba Fault system. Furthermore, the detailed regional Arabian Gulf–Gulf of Oman tectonic study carried out by Searle (159) for the Musandam Peninsula does not indicate any faulting system in the UAE south and west of the Dhaid–Ras Al Khaimah axis. This would appear to cast serious doubts on the existence of the WCF.
30.0 N
27.5 N
Dubai
25.0 N
22.5 N
20.0 N 50 .0 E
52.5 E
55.0 E
57.5 E
60.0 E
Figure A26 - Seismicity catalogue (Sigbjornsson and Elnashai (157) ).
The WCF (see Figure A27) was included in an estimate of damage caused by hypothetical seismic events in the UAE by Wyss and Al-Homoud(160), but the authors stress that it is not known whether or not the WCF is active. The inclusion of the WCF in an estimate of damage due to hypothetical seismic events is reasonable, but its inclusion in the Sigbjornsson and Elnashai seismic hazard assessment is contentious because there is neither apparent associated seismicity nor geological evidence of recent (Quaternary) seismic activity. The hazard spectra derived by Sigbjornsson and Elnashai are shown in Figure A28. These show hazard levels that are significantly higher than those derived in the site-specific assessments carried out by Bommer (149), Wood and Irvine (152) and Al-Homoud (161). The Sigbjornsson and Elnashai hazard levels are also significantly higher than those presented by Abdalla and Al-Homoud (150), and they are conservative and were later reduced by Al-Homoud.
15
Guide to the Design of Concrete Structures in the Arabian Peninsula
Figure A27 - Epicentres (black dots) for the UAE. Squares show cities. Solid grey lines show faults. Solid stars mark hypothetical epicentres and the white star marks the M w 5, 2002, epicentre (Wyss and Al-Homoud (160) ).
Figure A28 - Site-specific hazard spectra for Dubai for return periods of (A) 2475 years and (B) 947 years (Sigbjornsson and Elnashai (157) ).
A4.4
The 2475-year return period seismic hazard derived by Sigbjornsson and Elnashai (157) for Dubai is shown in Figure A29. Their spectrum is compared with the 2475-year spectra derived by Wood and Irvine (152) and Peiris et al.(116). It is evident that the Sigbjornsson and Elnashai hazard is approximately three and five times that of Wood and Irvine and Peiris et al. respectively.
Discussion of the site-specific seismic hazard assessments
The 2003 Al-Homoud(151) assessment has been shown to be deficient in that it used an unreasonably high earthquake in the Northern Emirates, which led to an overestimate of the seismic hazard. This error was also present in the Abdalla and Al-Homoud study (150) , but was later (in 2004) corrected by AlHomoud (156) .
Figure A29 - 2475-year return period seismic hazard spectra for Dubai.
16
Appendix A – Seismic design
Comparison with the hazard spectrum for an Mw 6.5 event at 20km suggests that the Sigbjornsson and Elnashai hazard is strongly influenced by high-magnitude near-source events, probably the inferred West Cost Fault (WCF). It has been stated earlier that the inclusion of the WCF in a seismic hazard assessment is contentious because it is not known if the fault exists. Furthermore, there is neither apparent associated seismicity nor geological evidence of recent (Quaternary) s eismic activity. In addition, a study by Aldama-Bustos et al , (A1) compares favourably with the low seismic hazard presented by Peiris et al (116) and Musson et al (117) assessments. Therefore, the Sigbjornsson and Elnashai seismic hazard assess-ment must be considered to be very conservative.
A5.2
Comparison of PGA estimates
The estimates of PGA for Dubai from the site-specific and regional hazard assessments are shown in Table A2. Seismic assessment by: Wood and Irvine(152):
PGA (g) Reverse faulting Strike-slip faulting Weighted average
Al-Homoud(156) Peiris et al.
(116)
Musson et al.
(117)
McFarlane (153) Malkawi et al.
(112)
Figure A19
A5 A5.1
Discussion of the seismic hazard for the Arabian Gulf states Introduction
From the material presented in the previous sections of this appendix, the hazard for the southern Arabian Gulf states is generally low. Nevertheless, four studies suggest a higher hazard level and these have been discounted for the following reasons:
The assessment by Mortimer-Lloyd(144) has been included for historical completeness and it has been superseded by later studies. The Grunthal et al.(111) study has no scientific basis for the southern Arabian Gulf states. The Abdalla and Al-Homoud (150) study unjustifiably extends high seismicity from the Zagros region significantly southwards into the UAE. Sigbjornsson and Elnashai(157) obtain a high hazard by including the inferred WCF as a major seismic source, but there is neither apparent associated seismicity nor geological evidence of historical seismic activity. (110)
The Al-Haddad et al. hazard assessment has been included in the Saudi Arabian building regulations and therefore may be used for most of the Arabian Peninsula. However, it should be amended to include the later studies that have been carried out for the southern Arabian Gulf states, which are described below. Eight seismic assessments have been presented which can be considered to be reliable. These present a relatively low seismic hazard for the southern Arabian Gulf states. Five of the assessments are regional studies and three are site-specific for Dubai. The regional assessments are by Sadek (154), Peiris et al.(116), McFarlane(153) , Musson et al.(117) and Malkawi et al.(112) . The site-specific assessments were carried out for Dubai by Bommer (149), Wood and Irvine (152) and Al-Homoud (156) . These assessments (regional and site-specific) are used as the basis for discussion of seismic hazard.
0.078 0.058 0.072 0.075 0.060 0.050 0.050 0.075 0.072
Table A2 - Comparison of PGA estimates.
The Musson et al.(117) and McFarlane (153) assessments estimate PGA for Dubai to be 0.05g, whereas the other assessments estimate PGA to be within the 0.058g to 0.075g range. This range of PGA estimates is not unreasonable and is probably due to the differences in the attenuation relationship. For example, Wood and Irvine (152) present a weighted average for PGA for the Abrahamson and De Silva (162) attenuation relationship. This average is based on the assumption that ground motions in Dubai from reverse faulting are twice as likely as ground motions from strike-slip faulting. Their PGA estimates for reverse faulting and strike-slip faulting are 0.058g and 0.078g respectively – a not insignificant difference for a variation of the same attenuation relationship. Furthermore, Figure A31a indicates that the Abrahamson and De Silva relationship predicts higher PGA than the Ambraseys and Melville(143) relationship for Mw 5.5 events up to a distance of approximately 150km. The former relationship was used by Wood and Irvine and Al-Homoud and the latter by Peiris et al. and Musson et al. From the foregoing, the PGA estimates for Dubai are in reasonable agreement and range from 0.05g to 0.075g. The differences in the PGA estimates would appear to be due to the choice of attenuation relationship. The 950-year return period PGA of 0.125g to 0.15g derived by Sadek (154) for Kuwait City compare reasonably well with the PGA values of 0.12g and 0.27g derived by Peiris et al. for return periods of 475 years and 2475 years respectively. Consequently, either estimate may be used for design, but the Peiris et al. values are based on more common return periods and therefore are more useful for design.
17
Guide to the Design of Concrete Structures in the Arabian Peninsula
A5.3
Comparison of seismic hazard spectra
Peiris et al.(116) derived seismic hazard spectra for several major cities in the Arabian Gulf Region and Musson et al.(117) derived seismic hazard spectra for seven major cities in the UAE. The hazard derived by Peiris et al. for the cities is similar, with the exception of Kuwait City, which has a higher seismic hazard (see Table A1). Furthermore, Peiris et al. provide greater details regarding the hazard in Dubai and consequently their data for Dubai is used to discuss matters relating to the regional seismic hazard. The seismic hazard spectra for Dubai from the pertinent sitespecific and regional hazard assessments are shown in Figure A30. It is evident that the UBC 97 Zone 1 spectrum for soft rock sites provides a reasonable estimate of the Wood and I rvine (152) and Al-Homoud(156) spectra. It is also evident that the UBC 97 Zone 1 spectrum for rock sites provides a reasonable envelope of the Peiris et al. and Musson et al. spectra at periods greater than 1s. Nevertheless, despite the PGA estimates derived by Peiris et al. and Musson et al. being in reasonable agreement with those derived by Wood and Irvine and Al-Homoud, the spectral accelerations at periods greater than zero are not in good agreement. Part of the discrepancy can be explained by the Peiris et al. and Musson et al. spectra being based on bedrock values (shear wave velocity, V s > 760m/s), whereas the Wood and Irvine and Al-Homoud spectra are based on soft rock values (shear wave velocity, 360m/s < V s < 760m/s). This would reduce the PGA estimates by approximately 20% and the short-period spectral ordinates even further.
Figure A30 - Comparison between UBC 97 and Dubai site-specific spectra.
18
Furthermore, another difference in the response spectra is probably due to the attenuation relationships used. Peiris et al. used different relationships for different source regions and they used the attenuation equations by Ambraseys and Melville (143) and Sadigh et al.(163) for the Zagros region and the Makran region. Musson et al. also used the Ambraseys and Melville relationship, whereas Wood and Irvine and Al-Homoud used the Abrahamson and De Silva (162) relationship. Wood and Irvine explain that they used the Abrahamson and De Silva relationship because this relationship is based on strong motion date recorded out to 300km. They note that the four attenuation relationships considered by them show reasonable agreement for distances up to 60k m. They also note that the spectral accelerations in the 100–300km range are significantly higher for the Abrahamson and De Silva relationship when compared with the other relationships in general and the Sadigh et al. relationship in particular. Therefore, it would be informative to compare the Abrahamson and De Silva attenuation relationship used by Wood and Irvine and Al-Homoud with the Ambraseys and Melville relationship used by Peiris et al. and Musson et al. Figure A31 shows the PGA and short-period (0.2s) spectral acceleration for an Mw 5.5 earthquake. This magnitude of earthquake is only likely to have structural implications for near-source events. It is evident from Figure A31 that the Abrahamson and De Silva relationship predicts higher short-period accelerations for distances up to 120km. Furthermore, at 30km and 50km the Abrahamson and De Silva 0.2s spectral accelerations are 50% and 30% greater than the corresponding Ambraseys and Melville values.
Appendix A – Seismic design
(a) PGA
(b) Spectral acceleration at T = 0.2s
Figure A31 - Attenuation of PGA and short-period (0.2s) acceleration at rock sites for an M w 5.5 earthquake at reverse fault sources.
(a) Spectral acceleration at T = 1s
(b) Spectral acceleration at T = 2s
Figure A32 - Attenuation of long-period acceleration at rock sites for an M w 7 earthquake at reverse fault sources.
Figure A32 shows the long-period (1 and 2s) spectral acceleration for an Mw 7 earthquake. An Mw 7 earthquake is likely to have structural implications for distant-source events. It is evident that the Abrahamson and De Silva relationship predicts higher long-period accelerations for distances greater than about 100km. Furthermore, for distances of 200– 300km the Abrahamson and De Silva spectral accelerations are 40–80% greater than the corresponding Ambraseys and Melville values.
From the foregoing, it is evident that the Abrahamson and De Silva attenuation relationship tends to predict higher spectral accelerations than the Ambraseys and Melville relationship for the small-magnitude near-source events and large-magnitude distant-source events pertaining to the southern Arabian Gulf states. This probably explains the higher s pectral accelerations derived by the Wood and Irvine and Al-Homoud response spectra when compared with those derived by Peiris et al. and Musson et al., as shown in the response spectra presented in Figure A30.
19
Guide to the Design of Concrete Structures in the Arabian Peninsula
Therefore, until generally agreed attenuation relationships are available for the Arabian Gulf Region, it would be prudent to include the Abrahamson and De Silva (162) relationship in seismic hazard assessments. It would also be prudent to give it more weight than other relationships that are generally in use.
A5.4
Site response effects
Seismic hazard in the southern shores of the Arabian Gulf Region is due to either small-magnitude local earthquakes that are not associated with known tectonic features or large-magnitude distant events that are associated with known tectonic features. Peiris et al.(116) modelled these site response scenarios by scaling their bedrock motions for Dubai to the October 2001, Melton Mowbray, UK, Earthquake (Ms 3.9 at 15km) and the September 1985 Mexico City Earthquake (Mw 8.1 at 400km). The Melton Mowbray record was selected to represent a low-magnitude nearsource earthquake scenario and the Mexico City record was selected to represent a large-magnitude distant-source earthquake scenario. The Peiris et al. spectra are shown in Figure A33. Figure A33a shows the very dense soil response spectra. There is a significant amplification of the ground surface response for
the near-source (Melton Mowbray) earthquake scenario in the short-period range up to 0.3s. Conversely, the distant-source (Mexico City) earthquake scenario shows the ground surface response spectrum to be similar to the bedrock spectrum. Figure A33b shows the soft soil response spectra. The nearsource (Melton Mowbray) earthquake scenario was not amplified in the short-period range. However, the distant-source (Mexico City) earthquake scenario shows significant amplification of the ground surface response spectrum in the medium to long period range. The Peiris et al. PGA of 0.6g and bedrock spectrum shown in Figure A33 would place Dubai and most of the other Arabian Gulf cities in UBC 97 Zone 0, for which no seismic design is required. Nevertheless, it is proposed to allow for the near-source and distant-source earthquake scenarios by specifying a minimum Arabian Gulf Earthquake, Zone Z AG (the suffix AG meaning Arabian Gulf). Eurocode 8 specifies a minimum reference PGA of 0.05g for seismic design. Therefore, it is reasonable to base the Zone ZAG spectrum on a PGA of 0.05g. The spectrum is derived by linear extrapolation of the UBC 97 coefficients, C a and C v, as shown in Table A3. These coefficients can also be used to derive appropriate International Building Code (IBC) and Eurocode 8 coefficients as shown in the design section of this report.
(a) Very dense soil and stiff rock spectra, UBC 97 soil type S C
(b) Soft soil spectra, UBC 97 soil type S E Figure A33 - Comparison of 475-year return period bedrock and ground sur face spectra for Dubai, 5% damping (Peiris et al. (116) ).
20
Appendix A – Seismic design
Spectrum classification
Seismic coefficient
Soil profile type SA
SB
SC
SD
SD
UBC 97 Zone 1 (0.075g PGA)
C a
0.060
0.075
0.090
0.120
0.190
C v
0.060
0.075
0.125
0.180
0.260
Interpolated (0.060g PGA)
C a
0.048
0.060
0.072
0.100
0.168
C v
0.048
0.060
0.100
0.152
0.212
Zone ZAG (0.05g PGA)
C a
0.040
0.050
0.060
0.087
0.153
C v
0.040
0.050
0.083
0.133
0.180
Table A3 - Seismic coefficients C a and C v , for UBC 97 Zone 1, the proposed minimum Arabian Gulf Zone, Z AG , and the interpolated 0.06g PGA design spectra.
The applicability of the Zone ZAG spectrum is checked by comparison with the near-source and distant-source scenarios suggested by Peiris et al. These are summarised in Figure A34a to c. The Peiris et al. bedrock spectrum is included in each of these figures to allow scale comparison. Furthermore, in order to allow a correct comparison with the Peiris et al. data, interpolated spectra based on the Peiris et al. PGA of 0.06g are used for the comparison. The seismic coefficients for the interpolated spectra are shown in Table A3. These coefficients are based on linear interpolation of the UBC 97 Zone 1 and the proposed Zone Z AG coefficients for each of the soil profile types. Figure A34a compares bedrock response spectra for UBC 97 soil type SB and it shows that the near-source (Melton Mowbray) earthquake scenario exhibits large amplification at short period. However, it is evident that the interpolated spectrum provides a good estimate of this short-period amplification and that it also provides an excellent correlation with the equivalent Peiris et al. spectrum at long period. Figure A34a also shows that the Zone ZAG spectrum provides a safe envelope for the equivalent Musson et al.(117) spectrum at short period and a slight underestimate at long period. Figure A34b compares surface response spectra for UBC 97 soil type SD. It should be noted that Peiris et al. used UBC 97 soil type SC for their calculated average shear wave velocity , V s = 455m/s. This is technically correct; however, UBC 97 specifies soil type SD for V s = 360m/s, which is closer to 455m/s than the 760m/s upper range for soil type S C. Therefore, soil type SD is used for the comparison in Figure A34b. The near-source (Melton Mowbray) earthquake scenario exhibits large amplification at short period. However, it is evident that the interpolated spectrum provides a good estimate of this amplification Figure A34c compares surface response spectra for UBC 97 soil type SE. The distant-source (Mexico City) earthquake scenario exhibits large amplification of the ground surface response in the medium- to long-period range. However, it is evident that the Zone Z AG spectrum and the interpolated spectrum provide a safe envelope of this amplification.
The Eurocode 8 and IBC spectra were also compared with the earthquake scenarios shown in Figure A34, by using the spectral accelerations described in Table A1. The Eurocode 8 spectra give similar estimates of the ground surface amplifications to those derived by the interpolated spectra for all periods. However, the IBC spectra underestimate the amplifications at short period and overestimate them at long period. From the foregoing, it is evident that the proposed Zone Z AG classification allows a very good correlation with the near-source and distant-source ground surface amplification scenarios suggested by Peiris et al. Therefore it would be prudent to use the Zone ZAG classification to replace UBC 97 Zone 0 in areas of low seismicity in the Arabian Gulf Region. This would allow for site response amplification effects that are not modelled by Zone 0 but which are particular to the region, namely small-magnitude near-source and large-magnitude distant-source earthquake scenarios. Linear interpolation between the UBC 97 Zone 1 and Zone ZAG coefficients may be used where appropriate.
A5.5
Proposed seismic hazard for the Arabian Gulf Region
The PGA isoseismals from the probabilistic seismic assessments shown in Figures A20 and A23 (Musson et al.(117) and Malkawi et al.(112) respectively) show a similar shape of seismic hazard. However, the Malkawi et al. hazard assessment, with the isoseismals extending further southwards, is generally more conservative than the Musson et al. assessment. The only major exception to this is that Musson et al. show a higher hazard between Dibba and Fujeirah on the east coast of the UAE. This is probably due to their more detailed study of the seismotectonics of this area. Furthermore, the shape and distribution of the seismic hazard derived by these probabilistic assessments is consistent with the deterministic assessments presented in this appendix. The McFarlane (153) assessment shown in Figure A18 shows a good correlation with the Musson et al. assessment, while Figure A19 depicts a similar agreement with the Malkawi et al. assessment.
21
Guide to the Design of Concrete Structures in the Arabian Peninsula
(a) Melton Mowbray, UBC 97 and other bedrock spectra, soil type S B
(b) Melton Mowbray and UBC 97 stiff soil spectra, soil type S D
(c) Mexico City and UBC 97 soft soil spectra, soil type S E Figure A34 - Comparison between UBC 97 and 475-year return period site response spectra for 5% damping (after Peiris et al. (116) ).
22
Appendix A – Seismic design
Country
City
Peiris et al.(116) (g)
Musson et al.(117) (g)
Malkawi et al.(112) (g)
Proposed minimum design PGA (g)
Bahrain
Manama
0.060
–
–
0.060
Kuwait
Kuwait City
0.120
–
–
0.120
Oman
Muscat
0.050
–
–
0.050
Qatar
Doha
0.040
–
0.030
0.050
Saudi Arabia
Dammam
–
–
–
0.050
Al Khobar
–
–
–
0.050
Abu Dhabi
0.060
0.035
0.060
0.060
Dubai
0.060
0.050
0.075
0.075
Fujeirah
–
0.080
0.065
0.080
Umm Al Quain
–
0.065
0.080
0.080
Ras Al Khaimah
–
0.800
0.100
0.100
UAE
Table A4 - Comparison of PGA estimates for cities in the Arabian Gulf region.
From the foregoing, it would be reasonable to consider the seismic hazard maps prepared by Malkawi et al. and Musson et al. as upper and lower bounds respectively for the seismic hazard in the Arabian Gulf Region. However, for design purposes, it would be prudent to supplement these hazard maps by using the conservative design assumptions discussed in the following sections.
A5.6
Proposed minimum design PGA
The Peiris et al.(116) and the Musson et al.(117) estimates of PGA are shown in Table A4. These are compared with the estimates of PGA derived from the Malkawi et al.(112) isoseismals shown in Figure A23 and the proposed minimum Zone Z AG classification (PGA of 0.05g) for areas of low seismicity in the Arabian Gulf Region. Inspection of Table A4 indicates that the proposed minimum design values provide conservative, but not excessively high, estimates of PGA.
A6 Conclusions and recommendations A6.1
Summary
Several seismic hazard assessments have been presented and it is generally shown that the seismic hazard in the southern states of the Arabian Gulf is relatively low. Those assessments that show a higher seismic hazard have been disregarded as having a weak scientific basis, with the exception of the study by Sigbjornsson and Elnashai (157) which is discussed below.
A6.2
Implications of the inferred UAE West Coast Fault
Sigbjornsson and Elnashai(157) obtain a high hazard by including the inferred UAE West Coast Fault (WCF) as a major seismic source. However, there is neither apparent associated seismicity nor geological evidence of historical seismic activity to prove the existence of the WCF. Therefore, its inclusion in the Sigbjornsson and Elnashai seismic hazard assessment is difficult to justify. The WCF is shown in the tectonic map of Saudi Arabia (see Johnson(142) ) and the source for the fault trace on the map appears to be Murris (158). However, there is no direct reference to the WCF in Murris, but he does interpret it to be part of the Dibba Fault system. Furthermore, the regional tectonic study carried out by Searle (159) does not indicate any faulting system in the UAE south and west of the Dhaid–Ras Al Khaimah axis. This would appear to cast serious doubt on the existence of the WCF. The inclusion of such a fault structure in a seismic assessment should be based on a detailed seismotectonic study being carried out to prove its existence. This should include:
a review of oil industry seismic and well data geomorphic mapping paleoseismic studies.
If the existence of the WCF were to be proved by such a study, and the degree of seismicity established, then it should be included in seismic hazard assessments. This could significantly increase the seismic hazard along the entire west coast of the UAE. For example, the 2475-year return period spectrum presented by Sigbjornsson and Elnashai (157) scales to a 475-year spectrum that is equivalent to UBC 97 Zone 4. This seismic hazard level appears to be excessively high and it equates to the worst areas of California. This high hazard level would run for nearly the entire length of the UAE and would extend for several tens of kilometres inland from its western coastline.
23
Guide to the Design of Concrete Structures in the Arabian Peninsula
A detailed discussion of the hazard posed by the inferred WCF is presented by Aldama-Bustos et al (A1) and we concur that including the WCF in probabilistic assessments would be excessively conservative.
A6.3
Seismic hazard for the southern Arabian Gulf Region
The regional probabilistic seismic assessments presented in this appendix that are deemed to be reliable show a low seismic hazard for the southern Arabian Gulf Region. The probabilistic assessments are reasonably consistent with their level of hazard and these hazard levels are corroborated by two regional deterministic assessments. Furthermore, there is a reasonable correlation between the PGA estimates derived from the regional assessments and those that are derived from the site-specific assessments for Dubai. Yet, these assessments use different methodology, source zones and attenuation relationships. This indicates that their prediction of seismic risk may be treated with a high degree of confidence. It is suggested that Figures A23 and A20 may be treated as upper and lower bounds of the seismic hazard in the Arabian Gulf Region. It is concluded that Figure A23 may be used for seismic design, if supplemented by the minimum design PGA shown in Table A4. Furthermore, a minimum Zone Z AG classification (PGA of 0.05g) for areas of low seismicity in the Arabian Gulf Region is shown to be more appropriate than UBC 97 Zone 0. However, although there is a reasonable correlation between the PGA estimates from the various assessments, there is not a good agreement between their hazard response spectra. It is shown that the major difference in the response spectra is due to the attenuation relationships used and their differing definition of shear wave velocity at bedrock.
It is apparent that the Abrahamson and De Silva (162) attenuation relationship produces higher spectral accelerations for the particular conditions pertaining to the southern Arabian Gulf Region, namely small-magnitude near-source and large-magnitude distant-source earthquake scenarios. Therefore, until generally agreed attenuation relationships are available for the region, it is recommended that the Abrahamson and De Silva relationship be included in seismic hazard assessments. Consideration should also be given to assigning a higher weighting to this relationship than other relationships generally in use.
A6.4
Recommended seismic hazard for the Arabian Peninsula
It is recommended that the Al-Haddad et al.(110) hazard map shown in Figure A6 may be used for seismic design in most of the Arabian Peninsula, but it should be amended to include the seismic hazard data described below:
The Malkawi et al.(112) hazard map shown in Figure A23 may be used for seismic design in the southern Arabian Gulf Region provided it is supplemented by the minimum design requirements shown in Table A4 and summarised as follows. The minimum seismic hazard in the southern Arabian Gulf Region should be classified as Zone Z AG, which has a minimum PGA of 0.05g. The minimum seismic hazard response spectrum should be based on the Zone ZAG classification. Adaptations of the UBC 97 seismic coefficients, C s and C v, for Zone Z AG spectra are presented in Table A3. Linear interpolation between the Zone 1 and ZAG coefficients may be used where appropriate. The resulting coefficients, C s and C v, may also be used to derive appropriate IBC and Eurocode 8 coefficients as shown in the design section of this report.
The above recommendations represent a conservative estimate of the seismic hazard for the Arabian Peninsula, based on current evidence. However, the use of alternative seismic data is not precluded. Examples of less conservative hazard levels are presented by Peiris et al.(116) Musson et al.(117) and Aldama-Bustos et al (A1). An example of a significantly higher hazard level is that presented by Sigbjorn-sson and Elnashai (157) .
Additional reference A1. ALDAMA-BUSTOS, G, BOMMER, JJ, FENTON, CH AND STAFFORD, PJ, Probabilistic seismic hazard analysis for rock sites in the cities of Abu Dhabi, Dubai and Ra’s Al Khaymah, United Arab Emirates, Georisk: Assessment and Management of Risk for Engineered Systems and Geohazards, pp 1-29, Vol 2, September 2008
24
Appendix B – Wind design
Appendix B – Wind design B1
Overview of wind climate in the Arabian Gulf
The wind climate of the Arabian Gulf described in this appendix is based on research carried out by Davids (118,119). The structure and organisation of natural wind over the region is characterised by four different types of air movement: 1. 2. 3. 4.
synoptic winds over a large scale thunderstorm downburst winds over a small scale unique regional winds, called shamals cyclonic events originating in the north-western Indian Ocean. Figure B1 - Radar image of large-scale air mass.
Each of these types of systems produces strong winds near ground level, but their effects upon buildings are quite different. This is due to their different characteristics of scale, profile energy, directionality and duration, as discussed below.
B2
Characteristics of Arabian Gulf wind systems
A typical synoptic velocity profile is s hown in Figure B2, which shows the increase in wind speed with height within the boundary layer. This variation of mean speed with height is usually described by various power law profiles. 1000
800
B2.1
Synoptic winds
Description Synoptic winds are produced by large-scale air masses which move from areas of high pressure to areas of low pressure.
) m ( t h g i e H
600
400
200
Scale Synoptic winds are typically 1000km or more in extent. For example, a radar image of a typical large-scale air mass moving over the south-eastern United States is shown in Figure B1. This image shows an intense low-pressure cell centred on northern Florida, with associated strong winds and cloud arcing on a front all the way into Canada. Profile A large mass of air flowing over the earth’s surface is slowed down and made turbulent by the roughness of the surface. This turbulent layer is known as the boundary layer. As the distance from the surface increases, the surface roughness has less effect and wind speed slowly increases with height until a height is reached where the influence of the surface roughness is negligible. This height is referred to as the gradient height.
0 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Vmean /Vmean (600m)
Figure B2 - Velocity profile of synoptic winds (from Aurelius (164) ).
Direction Winds blow from areas of high pressure to areas of low pressure. The pre-dominant direction is from W–NNW. However, winds in coastal areas tend to exhibit a diurnal pattern, with onshore winds during daylight hours changing to offshore at night. Refer to the design section of this report for more information on wind direction in the Arabian Peninsula. Energy Low to high.
25
Guide to the Design of Concrete Structures in the Arabian Peninsula
Duration
Profile
Ranging from diurnal variations to durations of s everal days.
A typical velocity profile of a thunderstorm downburst is shown in Figure B3c.
Frequency Direction Many times per year; the predominant wind system in the Arabian Gulf.
B2.2
Direction is 360° from point of impact of downburst; therefore any direction. Energy
Thunderstorm downbursts
Description
Medium to high.
A well-developed thunderstorm cloud is shown in Figure B3a. Such clouds typically have a cloud base at a height of 1000– 2000m, and rise to perhaps 5000m. Such thunder clouds are created by a column of warm air several hundred metres in diameter which rises and cools rapidly. This rapid cooling creates strong currents which circulate vertically within the cloud, further fuelling the process. These local cells quickly become unstable, and often collapse in the late afternoon, releasing a sudden burst of high-velocity air downwards as the cloud collapses through its own base.
Duration
Scale
Several minutes. Frequency Typically five or six per year. Thunderstorms contaminate the synoptic wind data set and therefore thunder-days should be separated from the synoptic data, otherwise higher estimates of mean wind speed result as shown in Figure B4.
Small-area local events.
(a) Typical thundercloud Figure B3 - Typical thunderstorm downburst.
26
(b) Computer simulation
(c) Profile (from Mason et al.(165))
Appendix B – Wind design
Figure B4 - Separation of synoptic and thunderstorm data (from Aurelius (164) ).
B2.3
Arabian Peninsula shamals
Description The wind climate in the Arabian Gulf includes a unique regional flow called the shamal , which is the Arabic word for northern. These winds tend to occur in late summer as a result of temperature inversions over the Arabian Desert. They come from a north-westerly direction and travel down the Arabian Gulf. Satellite images of the Gulf during a normal day and during a shamal are shown in Figure B5.
Profile The velocity profiles of typical shamals are shown in Figure B6. The peak velocity tends to occur at heights of approximately 150–400m. The velocity profile of a shamal is significantly different from the synoptic profile and consequently a typical anemometer at 10m height would not be a good indicator of the shamal strength at height. Direction Typically from the north-west.
Scale Energy Large-scale Arabian Gulf-wide events. Medium to high.
(a) Normal conditions
(b) During shamal
Figure B5 - Satellite image of shamal.
27
Guide to the Design of Concrete Structures in the Arabian Peninsula
Figure B6 - Shamal velocity profile (from Membery (166) ).
Duration
Scale
Usually several days, but can last up to 40 days.
The size of a cyclone is usually determined by measuring the distance from the centre to the outermost closed isobar. Very small cyclones have a radius less than 200km, but cyclones typically have a radius of 700–900km.
Frequency Typically one shamal season per year, in late summer.
B2.4 Cyclones (originating in north-western Indian Ocean) Description
Profile The wind speed of a cyclone is low at the eye; it is highest at the eye wall (see Figure B9a) and it decreases thereafter, with increasing distance from the eye. The wind speed at the eye wall is relatively constant with height (see Figure B9b). Direction
A typical well-developed cyclone is s hown in Figure B7a. A cyclone is a convective storm system which develops over a large area of warm ocean surface. The term cyclone is used to describe the cyclonic nature of the storm; c yclones rotate anticlockwise in the northern hemisphere and clockwise in the southern hemisphere. Depending on the geographical location, cyclones may also be called hurricanes or typhoons. Cyclones are characterised by a low-pressure centre (eye) which is surrounded by rain bands (see Figure B8). The primary energy source of the cyclone is the release of condensation at high altitude. The condensation leads to higher wind speeds because some of the released energy is converted into mechanical energy. The higher wind speeds and lower associated pressures lead to increased surface evaporation, which leads to more condensation. This provides the convective system with enough energy to be self-sufficient and maintain the positive feedback loop. However, to continue to maintain the positive feedback loop the cyclone must remain over warm water. When a cyclone passes over land, it is cut off from the heat source and its strength diminishes rapidly.
28
The movement of cyclones is dictated by large-scale (synoptic) winds. Cyclones generally originate on the eastern side of oceans and tend to move westwards, intensifying as they move (see Figure B7b). Energy Large-scale systems which can generate enormous energy. Ten-minute sustained wind speeds are greater than 34m/s for cyclones and estimated at 85m/s for severe cyclones. Duration Usually many days. Frequency Relatively infrequent and tend only to affect the eastern shores of the Arabian Peninsula. However, two cyclones made landfall in Oman during 2007.
Appendix B – Wind design
(a) Satellite image
(b) Route and strength
Figure B7 - Cyclone Gonu (June 2007).
Figure B8 - Cyclone convective storm system.
(a) Wind speed profile at 10m level (Typhoon York passing over Waglan Island, 1999)
(b) Vertical wind speed profile (Hurricane Erika, 1997)
Figure B9 - Cyclone wind speed profiles.
29
Guide to the Design of Concrete Structures in the Arabian Peninsula
B3
Conclusions
The main conclusion is that the wind climate in the Arabian Peninsula is a ‘mixed-climate’ with a particular emphasis on shamals, which are unique to the region. The climate is a complex mix of events that are generated by different physical conditions. It is evident that the strength, direction, duration, probability of occurrence and, in particular, the velocity profiles of the various wind events are significantly different.
30
Appendix C – Test methods for self-compacting concrete
Appendix C – Test methods for self-compacting concrete C1
Slump-flow test and T 500 test
The slump-flow test is used to assess the horizontal free flow of self-compacting concrete (SCC) in the absence of obstructions. It was first developed in Japan for use in assess ment of underwater concrete. The test method is based on the test method for determining the slump. The diameter of the concrete circle is a measure for the filling ability of the concrete. The apparatus consists of a mould in the shape of a truncated cone with the internal dimensions 200mm diameter at the base, 100mm diameter at the top and a height of 300mm, conforming to Part 2 of BS EN 12350, Testing fresh concrete(167). The base plate (see Figure C1) is formed from a stiff non-absorbing material, at least 700mm square, marked with a circle indicating the central location for the slump cone, and a further concentric circle of 500mm diameter. In addition a trowel, scoop, ruler and stopwatch (optional) are required.
Calculate the average of the two measured diameters. (This is the slump-flow in mm.)
C2
V-funnel test and V-funnel test at T 5 minutes
The V-funnel test, which is used to determine the filling ability (flowability) of the concrete with a maximum aggregate size of 20mm, was developed in Japan. The equipment consists of a Vshaped funnel, shown in Figure C2.
Figure C2 - V-funnel. Figure C1 - Base plate for Slump-flow test.
The test procedure is as follows:
About 6 litres of concrete is needed to perform the test, sampled normally. Moisten the base plate and inside of slump cone. Place base plate on level stable ground and the slump cone centrally on the base plate and hold down firmly. Fill the cone with the scoop. Do not tamp; simply strike off the concrete level with the top of the cone using the trowel. Remove any surplus concrete from around the base of the cone. Raise the cone vertically and allow the concrete to flow out freely. Simultaneously, start the stopwatch and record the time taken for the concrete to reach the 500mm spread circle. (This is the T 500 time.) Measure the final diameter of the concrete in two perpendicular directions.
The additional equipment required consists of a bucket (about 12 litres capacity), trowel, scoop and stopwatch. The procedure for the V-funnel test is as follows:
About 12 litres of concrete is needed to perform the test, sampled normally. Set the V-funnel on firm ground. Moisten the inside surfaces of the funnel. Keep the trap door open to allow any surplus water to drain. Close the trap door and place a bucket underneath. Fill the apparatus completely with concrete without compacting or tamping; simply strike off the concrete level with the top using the trowel. Open within 10s after filling the trap door and allow the concrete to flow out under gravity. Start the stopwatch when the trap door is opened, and record the time for the discharge to complete ( the flow time). This is taken to be when light is seen from above through the funnel. The whole test has to be performed within 5 minutes.
31
Guide to the Design of Concrete Structures in the Arabian Peninsula
The procedure for the V-funnel test at T 5 minutes is as follows:
Do not clean or moisten the inside surfaces of the funnel again. Close the trap door and refill the V-funnel immediately after measuring the flow time. Place a bucket underneath. Fill the apparatus completely with concrete without compacting or tapping, simply strike off the concrete level with the top using the trowel. Open the trap door 5 minutes after the second fill of the funnel and allow the concrete to flow out under gravity. Simultaneously start the stopwatch when the trap door is opened, and record the time for the discharge to complete (the flow time at T 5 minutes). This is taken to be when light is seen from above through the funnel.
The additional equipment required consists of a trowel, scoop and stopwatch. The procedure for the test is as follows: About 14 litres of concrete is needed to perform the test, sampled normally. Set the apparatus level on firm ground; ensure that the sliding gate can open freely and then close it. Moisten the inside surfaces of the apparatus; remove any surplus water. Fill the vertical section of the apparatus with the concrete sample. Leave it to stand for 1 minute. Lift the sliding gate and allow the concrete to flow out into the horizontal section. Simultaneously, start the stopwatch and record the times taken for the concrete to reach the 200mm and 400mm marks. When the concrete stops flowing, the distances ‘H 1’ and ‘H 2’ are measured. Calculate H 2 / H1 ; this is the blocking ratio.
C3
L-box test method
This test, which is based on a Japanese design for underwater concrete, assesses the flow of the concrete, and also the extent to which it is subject to blocking by reinforcement. The apparatus is shown in Figure C3.
The whole test has to be performed within 5 minutes.
C4
J-ring test
The J-ring test, developed at the University of Paisley, is used to determine the passing ability of the concrete. The apparatus is shown in Figure C4 and consists of the J-ring mounted on a base plate of a stiff non-absorbing material, at least 700mm square, marked with a circle showing the central location for the slump cone, and a further concentric circle of 500mm diameter. 100 0 0 3
200 300 d1
d2
Figure C4 - J-ring test apparatus (all dimensions in mm).
Figure C3 - L-box test apparatus.
32
Appendix C – Test methods for self-compacting concrete
The additional equipment required consists of a trowel, scoop and ruler. The procedure for the test is as follows:
About 6 litres of concrete is needed to perform the test, sampled normally. Moisten the base plate and inside of slump cone. Place base plate on level stable ground. Place the J-ring centrally on the base plate and the slumpcone centrally inside it and hold down firmly. Fill the cone with the scoop. Do not tamp, simply strike off the concrete level with the top of the cone using the trowel. Remove any surplus concrete from around the base of the cone. Raise the cone vertically and allow the concrete to flow out freely. Measure the final diameter of the concrete in two perpendicular directions. Calculate the average of the two measured diameters, in mm. Measure the difference in height between the concrete just inside the bars and that just outside the bars. Calculate the average of the difference in height at four locations, in mm. Note any border of mortar or cement paste without coarse aggregate at the edge of the pool of concrete.
33
Guide to the Design of Concrete Structures in the Arabian Peninsula
Appendix D – Cement, slag, fly ash and silica fume properties It is important that all ingredients used in concrete comply with a recognised standard and this is particularly important in the case of cements and similar materials. Requirements from a number of different national standards are given in the tables which follow. Tables D1 and D2 give the mechanical, physical and chemical requirements from BS EN 197-1 (8). The requirements are given in terms of characteristic values. Many of the property requirements depend on the cement strength class. Within each
Property Compressive strength:
Early
Strength class
Requirement
32.5N 32.5R, 42.5N 42.5R, 52.5N 52.5R
≥16.0MPa at 7 days ≥10.0MPa at 2 days ≥20.0MPa at 2 days ≥30.0MPa at 2 days
Standard (28 days) 32.5 N, 32.5R 42.5N, 42.5R 52.5N, 52.5R
strength class there are two early strength classes. The class with ordinary early strength class is indicated by N and the class with high early strength is indicated by R. Compositional and physical requirements for cement from ASTM C 150(10) are given in Tables D3 and D4. Requirements for ground granulated blastfurnace slag, fly ash and silica fume from European Standards are given in Tables D4 to D6.
32.5 to 52.5MPa 42.5 to 62.5MPa ≥52.5MPa
Initial setting time
32.5N, 32.5R 42.5N, 42.5R 52.5N, 52.5R
≥75min ≥60min ≥45min
Soundness
All
≤10mm
Table D1 - Mechanical and physical requirements for cement from BS EN 197-1 (8). Property
Strength class
Requirement (%)
Loss on ignition
All
≤5.0
Insoluble residue
All
≤5.0
Sulfate content
32.5N, 32.5R, 42.5N
≤3.5
42.5R, 52.5N, 52.5R
≤4.0
All
≤0.10
Chloride content
Table D2 - Chemical requirements for Portland cement (CEM I) from BS EN 197-1 (8). Property
Type I
II
V
Silicon dioxide (SiO2), min: %
–
20.0
–
Aluminium oxide (Al2O3), max: %
–
6.0
–
Ferric oxide (Fe2O3), max: %
–
6.0
–
Magnesium oxide (MgO), max: %
6.0
6.0
6.0
When C3A is 8% or less
3.0
3.0
2.3
When C3A is more than 8%
3.5
N/A
N/A
Loss on ignition, max: %
3.0
3.0
3.0
Sulphur trioxide (SO3), max: %
Insoluble residue, max: %
0.75
0.75
0.75
Tricalcium aluminate (C3A), max: %
–
8
5
Tetracalcium alumino ferrite plus twice tricalcium aluminate [C4AF + 2(C3A)]
–
–
25
Table D3 - Compositional requirements for common cement types from ASTM C 150 (10).
34
Appendix D – Cement, slag, pfa and silica fume properties
Property
Type I
II
V
12
12
12
Turbidimeter
160
160
160
Air permeability
280
280
280
0.80
0.80
0.80
3 days
12.0
10.0
8.0
7 days
19.0
17.0
15.0
28 days
–
–
21.0
Initial, min
60
60
60
Final, max
600
600
600
Time of setting, min
45
45
45
Time of setting, max
375
375
375
Air content of mortar, volume %, max 2
Fineness, specific surface m /kg, min
Autoclave expansion, max: % Compressive strength, MPa, min
Time of setting, min
Gilmore
Vicat
Table D4 - Physical requirements for common cement types from ASTM C 150
(10)
.
Property
Required value
Loss on ignition
3.0% max (corrected for oxidation of sulfide)
Chloride content
0.10% max
SO3
2.5% max
Sulfide
2.0% max
MgO
18.0% max
Fineness
275m2 /kg min
Activity index
45% min of Portland cement concrete strength at 7 days 70% min of Portland cement concrete strength at 28 days
Initial set
Not more than twice as long as test cement on its own
Table D5 - BS EN 15167 (49) requirements for ggbs. Property
Required value
Loss on ignition
Category A – 5.0% max
Chloride content
0.10% max
SO3
3.0% max
Free CaO
1.0% max – soundness test not required 2.5% max but soundness test also required
Total CaO
10.0% max
Fineness
Category S – less than 12% retained on 45μm sieve
Water requirement
95% max
Activity index
75% min of Portland cement concrete strength at 28 days 85% min of Portland cement concrete strength at 90 days
Soundness
10mm max
Particle density
Not more than 200kg/m3 deviation from declared value
Table D6 - BS EN 450 (47) requirements for fly ash.
35
Guide to the Design of Concrete Structures in the Arabian Peninsula
Property
Required value
Loss on ignition
4.0% max
Chloride content
0.30% max
SO3
2.0% max
SiO2
85% min
Elemental silicon
0.4% max
Free CaO
1.0% max
Specific surface
15.0m2 /g min; 35.0m2 /g max
Dry mass of slurry
Within ± 2% of declared value
Activity index
100% min of OPC concrete strength at 28 days
Table D7 - BS EN 13263 (69) requirements for silica fume.
36
Appendix E – Topography and climate
Appendix E – Topography and climate Note: Temperatures in the tables are shade temperatures. In the sun, these may be as much as 10–15°C higher in calm weather (possibly less in wind). During the hottest summer weather in almost all lowland areas of the Peninsula there is danger of heat exhaustion or heatstroke.
Month
E1
Saudi Arabia
Saudi Arabia forms the greater part of the Arabian Penin sula. It is a large country, most of which is hot desert. It is bordered on the north by Jordan and Iraq. To the east of Saudi Arabia there is a short coastline on the Arabian (Persian) Gulf but there are land borders with the Gulf states of Kuwait, Qatar, the United Arab Emirates and Oman. There is a long land border with Oman and in the south-west of the Peninsula it borders Yemen, to the north of which it has a long coastline on the Red Sea and the Gulf of Aqaba. Details of the climate in Jeddah, on the coast of the Red Sea, and in Riyadh, in the centre of the Peninsula, are given in Tables E1a and E1b.
Temperature (°C)
Relative humidity (%)
Precipitation
0800 hours
1400 hours
Minimum
Lowest recorded
Average monthly (mm)
Ave No. days with 1mm +
29
19
9
58
54
5
0.8
35
29
18
11
52
52
0
0.3
38
29
19
13
52
52
0
0.3
April
40
33
21
12
52
56
0
0.5
May
42
35
23
13
51
55
0
0
June
47
36
24
19
56
55
0
0
July
42
37
26
21
55
50
0
0
August
42
37
27
23
59
51
0
0
September
42
36
25
21
65
61
0
0
October
41
35
23
20
60
61
0
0
November
41
33
22
17
55
59
25
2
December
34
30
19
10
55
54
31
1
Highest recorded
Average daily Maximum
January
33
February March
Table El(a) - Climate in Saudi Arabia – Jeddah, 6m, 21° 28’N 39°WE, 5 years. Month
Temperature (°C)
Relative humidity (%)
Precipitation
0500 hours
1600 hours
Average monthly (mm)
Ave No. days with 1mm +
Highest recorded
Average daily Maximum
Minimum
Lowest recorded
January
30
21
8
–7
70
44
3
1
February
33
23
9
–2
63
37
20
1
March
38
28
13
1
65
36
23
3
April
40
32
18
2
64
34
25
4
May
43
38
22
15
51
31
10
1
June
45
42
25
19
47
31
0
0
July
45
42
26
19
33
19
0
0
August
44
42
24
17
35
19
0
0
September
44
39
22
17
42
24
0
0
October
38
34
16
10
47
25
0
0
November
34
29
13
2
60
33
0
0
December
31
21
9
0
75
52
0
0
Table E1(b) - Climate in Saudi Arabia – Riyadh, 590m, 24° 39’N 46° 42’E, 3 years.
37
Guide to the Design of Concrete Structures in the Arabian Peninsula
E2
Bahrain
E3
Bahrain consists of one large island and a number of smaller ones lying off the coast of Saudi Arabia and west of the Qatar Peninsula. Its climate is similar to that of the Gulf coast of Arabia but somewhat modified as Bahrain is an island. Humidity is high throughout the year except when hot, dry winds blow off the mainland. The high temperatures between April and October are rendered particularly uncomfortable by the humidity. Annual rainfall is low and this primarily falls between the months of November and March. Winter temperatures are mild and it is only rarely cold, when northerly winds blow from Iran.
Kuwait
Kuwait has land borders with Iraq and Saudi Arabia and a coastline on the Arabian Gulf. It is a low-lying hot desert country where the average annual rainfall is about 125mm. Most rain falls between November and March and there are very few wet days. Winter temperatures are mild and only occasionally is it cold, when northerly or north-westerly winds bring cold air from Iran or Iraq. Summers are uniformly hot and temperatures can soar when hot winds blow from the heart of Arabia. On the coast temperatures are lower than inland but the heat is made more uncomfortable by the high humidity. Occasional sandstorms occur when strong winds blow from the interior.
Details of the climate of Bahrain are given in Table E2. Details of the climate of Kuwait are given in Table E3.
Month
Temperature (°C)
Relative humidity (%)
Precipitation
0730 hours
1530 hours
Average monthly (mm)
Ave No. days with 2.5mm +
Highest recorded
Average daily Maximum
Minimum
Lowest recorded
January
29
20
14
5
85
71
8
1
February
34
21
15
7
83
70
18
2
March
35
24
17
11
80
70
13
1
April
41
29
21
13
75
66
8
1
May
42
33
26
19
71
63
0
0
June
44
36
28
21
69
64
0
0
July
44
37
29
24
69
67
0
0
August
45
38
29
24
74
65
0
0
September
44
36
27
22
75
64
0
0
October
39
32
24
19
80
66
0
0
November
36
28
21
14
80
70
18
1
December
31
22
16
9
85
77
18
2
Table E2 - Climate in Bahrain – 6m, 26° 129’N 50° 30’E, 16 years. Month
Temperature (°C)
Relative humidity (%)
Precipitation
Highest recorded
Average daily
1430 hours
Minimum
Lowest recorded
0530 hours
Maximum
Average monthly (mm)
Ave No. days with 1mm +
January
28
16
9
1
77
61
23
2
February
26
18
11
2
68
61
23
2
March
32
22
15
4
72
61
28
2
April
39
28
20
12
67
55
5
0.9
May
43
34
25
16
67
55
0
0.3
June
48
37
28
22
62
49
0
0
July
48
39
30
26
45
41
0
0
August
46
40
30
20
50
46
0
0
September
47
38
27
19
52
51
0
0
October
41
33
23
14
64
60
3
0
November
38
25
17
6
66
59
15
1
December
26
18
12
2
76
65
28
3
Table E3 - Climate in Kuwait – 5m, 29° 21’N 48° 00’E, 14 years.
38
Appendix E – Topography and climate
E4
Oman
E5
Oman is situated to the north-east of the Arabian Peninsula, with coastlines on the Gulf of Oman to the north and the Arabian Sea to the south. Inland it is bordered by the United Arab Emirates, Saudi Arabia and the Republic of Yemen. The Jebel Akhdar mountain range in the north of Oman rises to over 3000m where the annual rainfall on the higher parts exceeds 400mm. In the rest of Oman the annual rainfall is below 125mm except in the hills of Dhofar in the south. On the south coast the wet season is between June and September, but in the Jebel Akhdar and the lowlands of the north, rain may fall at any time. Occa sionally a tropical cyclone in the Arabian Sea brings very wet, windy weather to the coast of Oman which may cause damage through wind and flood. Climate details for Muscat are given in Table E4. Temperatures and humidity are high throughout the year on the coast; May to September is the hottest season. Temperatures rise inland towards the Rub al Khali, where humidity is lower so the high temperatures are more tolerable and the nights cooler.
United Arab Emirates
This territory consists of a union of seven sheikhdoms on the southern shore of the Arabian Gulf between Qatar on the west and Oman and the Gulf of Oman on the east. They have a boundary with Saudi Arabia to the north of the Rub al Khali. Much of the country is flat and consists of sand or rocky desert. Rainfall is very low and mostly occurs between November and March. Temperatures are very high between May and September and warm to mild for the rest of the year. Winters are warmer than in Kuwait or the interior of Saudi Arabia. Humidity is high on the coast in the summer. Table E5 for Sharjah is representative of conditions on the coast. Conditions only a few miles inland can be significantly less humid than at the coast. Hourly readings from Dubai International Airport during 2005 indicate an average relative humidity of 55% over the year compared with an average for Sharjah coast of 67% from the results given in Table E5.
Unless precautions are taken there is a danger of heat exhaustion or even heatstroke during the hottest weather. Sunshine amounts are high all year. Month
Temperature (°C)
Relative humidity (%)
Precipitation
0800 hours
1600 hours
Average monthly (mm)
Ave No. days with 0.25mm +
Highest recorded
Average daily Maximum
Minimum
Lowest recorded
January
31
25
19
11
72
71
28
2
February
32
25
19
12
73
73
18
1
March
42
28
22
17
71
70
10
1
April
41
32
26
19
64
68
10
1
May
44
37
30
24
58
60
0
0
June
47
38
31
26
72
72
3
0
July
45
36
31
25
77
77
0
0
August
42
33
29
24
82
80
0
0
September
42
34
28
23
75
77
0
0
October
41
34
27
21
69
74
3
0
November
36
30
23
17
69
72
10
1
December
33
20
20
16
70
71
18
2
Table E4 - Climate in Muscat – 5m, 23° 37’N 58° 35’E, 24 years. Month
Temperature (°C)
Relative humidity (%)
Precipitation
0730 hours
1530 hours
Average monthly (mm)
Ave No. days with 2.5mm +
Highest recorded
Average daily Maximum
Minimum
Lowest recorded
January
29
23
12
3
81
61
23
2
February
33
24
14
8
81
63
23
2
March
40
27
16
8
74
61
10
1
April
39
30
18
12
66
63
5
0.3
May
43
34
22
16
61
63
0
0
June
44
36
25
19
64
65
0
0
July
47
38
28
23
64
64
0
0
August
48
39
28
23
66
64
0
0
September
45
37
25
21
73
64
0
0
October
40
33
22
18
77
62
0
0
November
36
31
18
12
78
59
10
0.2
December
31
26
14
8
82
62
36
2
Table E5 - Climate in Sharjah – 5.5m, 25° 20’N 55° 24’E, 11 years.
39
Guide to the Design of Concrete Structures in the Arabian Peninsula
E6
The Yemen
The Yemen occupies the south-western portion and the extreme south of the Arabian Peninsula. In the south-western corner of the Peninsula it has a narrow coastal plain on the Red Sea with the land rising steeply to a mountainous interior over 3600m high. This is an exceptional part of Arabia as the mountains receive moderate to abundant rainfall between March and September. In the higher regions temperatures are much lower than elsewhere in Arabia. Here the climate is quite pleasant with mild winters and warm, moist but generally sunny summers. No reliable climatic data are available for this, higher, part of the country. In the lowland along the Red Sea coast the climate is hot and humid for most of the year and similar to that of the rest of the Red Sea coast of Saudi Arabia. Conditions here are represented by Table E6 for Kamaran Island off the Red Sea coast of the Yemen and Table E1a for Jeddah in Saudi Arabia. In this lowland rainfall is low, averaging about 100mm a year and may occur in winter or summer. Away from the coast, in the mountains, much higher levels of rainfall in winter supply water for reservoirs and irrigation systems. On the southern edge of the Peninsula the Yemen has a long Month
coastline on the Arabian Sea. Here the interior includes a small portion of the great sand desert of the Rub al Khali which is mainly in Saudi Arabia. Between this desert and the coast are ranges of hills within which runs a broad valley, the Wadi Hadhramaut. This area receives rather more rainfall and is settled and more densely populated. Climatic conditions along the coast are represented by Table E7 for Khormaksar, Aden’s airport. Rainfall is low throughout the year and most of the coastal plain consists of hot desert. Temperatures and humidity are high throughout the year and between June and September midday temperatures regularly rise to near 38°C with high humidity. Daily sea breezes help to mitigate the heat on the coast. Inland in the hills, temperatures and humidity are a little lower. Here, rainfall is a little more plentiful and mostly falls between May and September.
E7
Qatar
No separate table for the climate of Qatar has been included, as details of its climate are very similar to those shown in the tables for Bahrain and the United Arab Emirates.
Temperature (°C)
Relative humidity (%)
Precipitation
0900 hours
1500 hours
Minimum
Lowest recorded
Average monthly (mm)
Ave No. days with 1mm +
28
23
19
79
69
5
0.6
32
28
23
19
77
65
5
0.9
34
30
25
21
75
65
3
0.6
April
37
32
26
23
74
61
3
0.3
May
39
35
28
23
70
56
3
0.2
June
40
36
29
24
67
55
0
0.1
July
41
37
29
22
63
52
13
2
August
39
36
29
22
67
55
18
1
September
40
36
29
23
71
58
3
0.5
October
39
34
28
23
67
57
3
0.5
November
34
31
26
20
74
63
10
0.8
December
32
28
24
20
77
68
23
2
Highest recorded
Average daily Maximum
January
31
February March
Table E6 - Climate in Kamaran Island – 6m, 15° 20’N 42° 37’E, 26 years. Month
Temperature (°C)
Relative humidity (%)
Precipitation
0300 hours
1500 hours
Average monthly (mm)
Ave No. days with 1mm +
Highest recorded
Average daily Maximum
Minimum
Lowest recorded
January
30
28
22
16
78
63
5
1
February
31
28
23
17
79
65
0
0.5
March
35
30
24
19
82
66
5
0.3
April
37
32
25
20
83
66
0
0
May
39
34
27
24
83
66
0
0
June
41
37
29
26
76
51
0
0
July
40
36
28
23
76
49
5
1
August
38
36
28
23
78
50
3
0.7
September
38
36
28
25
78
56
0
0.2
October
38
33
24
19
77
58
0
0.2
November
33
30
23
18
77
61
0
0.2
December
31
28
23
17
76
62
5
2
Table E7 - Climate in Aden, Khormaksar Airport – 7m, 12° 50’N, 45° 01’E, 6 years.
40
Appendix F – Permeability testing
Appendix F – Permeability testing F1
Permeability tests
F1.1
Introduction
The requirements for these tests were introduced into specifications in the mid-1980s in an effort to improve the durability of reinforced concrete structures. The thinking behind the introduction was that many deterioration processes involve the movement of liquid or gas through concrete. Hence if the concrete is less permeable or less absorptive, it should be more durable. The principal difficulty with the approach is that the conditions in the specified tests are not necessarily directly comparable with the environment to which the concrete will be exposed in service. A further difficulty is that there are few, if any, theoretical models of deterioration processes which use the properties which are measured under the specifications. Other than for the Nordtest 443 (168), neither is there any way of calculating the parameters required by the models from the measured values. On the other hand, it cannot be claimed that the specified values are based on experience as insufficient practical experience has been gained since the widespread introduction of these tests. This means that the values which have to be attained under the specifications are mainly speculative and there is no way of assessing whether they will or will not lead to a structure which attains the required design life. In addition, the same requirements are often set for all concrete on a project even when there are different exposure conditions for different elements. Even though the introduction of tests related to durability does not have a particularly sound technical basis, it has had the positive benefit that it has forced concrete producers into the more widespread use of cement combinations. Materials such as silica fume, ground granulated blastfurnace slag and fly ash have been shown to improve the durability of concrete in other parts of the world. The use of durability testing in specifications rather than a prescriptive approach would allow concrete producers the freedom to develop their own solutions to the problem of producing a concrete to meet a particular value of absorption or permeability. Despite the limitations of these tests, there has been a gradual tightening of specifications over the years in terms of the values to be achieved. Recent research has shown that some of the tests give variable results with poor repeatability. This has meant that concrete suppliers have had to aim at even more stringent targets to reduce the risk of non-compliant results during production testing. There is a real danger of overspecification and the unnecessary increases in cost which this entails.
The limitations of the tests in terms of accuracy and repeatability lead to the conclusion that in most cases they may not suitable as routine quality control measures. It would be preferable to specify these tests for use at mix development stage and try to link them to a more easily measured parameter (say strength) for use in routine production quality control. Basic precautions need to be included in specifications such as always testing at least three samples from the same batch and taking the average as the result. If any one value is markedly different to the other two, discard that value and take the average of the other two as the result. It is unrealistic to expect the results of routine tests during production to attain the same values as those obtained in tests for initial acceptance. Similarly, it is unrealistic to expect results from cored site samples to produce the same value as those for tests on samples undertaken in the laboratory. This is because of differences in curing, particularly for samples taken at or near the surface; differences in compaction; sedimentation associated with concrete depth; and temperature stresses and internal movements. Values specified for routine tests s hould be less onerous than values specified for initial acceptance testing. The tests which are commonly specified are as follows:
BS EN 12390, Testing hardened concrete, Part 8, Depth of penetration of water under pressure(169) BS 1881, Testing concrete, Part 208, Recommendations for the determination of the initial surface absorption of concrete(170) BS 1881 Part 122, Method for determination of water absorption(170) Rapid chloride permeability test to either ASTM C1202, Standard test method for electrical indication of concrete’s ability to resist chloride ion penetration(24) or AASHTO T-277 (25).
More recent additions are Nordtest NT Build 443 (168) and Nordtest NT Build 492(171). Values specified for routine tests should be set on a statistical basis. As an example, in the case of water penetration to Part 8 of BS EN 12390: 1.
2.
average of any four consecutive results not greater than 15mm and no single result greater than 20mm.
In both cases, a result is the average value from a set of three samples from the same batch of concrete.
41
Guide to the Design of Concrete Structures in the Arabian Peninsula
Each of the above tests was developed for a specific purpose (sometimes for research) and it is probably true to say that none was originally intended to be a quality control tool for production of concrete. Conditioning of specimens or concrete surfaces prior to test is an influencing factor as, for example, dry surfaces are much more absorbent than damp sur faces. Another important consideration is the age of the sample at test as the pore structure and hence also the absorption and permeability of concrete change with time. Much of what follows is based on the papers by Pocock and Corrans (22) and Krieg(23).
head. Test results show good reliability but initial moisture condition or the presence of curing membrane can have a significant effect on results. It is the only durability test which can be used as a non-destructive test at site and can be used relatively easily on both top and side sur faces of members. It measures the properties of the surface zone and hence can give an indication of increased absorption where curing has been deficient.
F1.5 F1.2
BS EN 12390 Part 8, Depth of penetration of water under pressure
In the test method given in BS EN 12390 Part 8, water under a pressure of 500kPa (5 bar) is applied to an area of one surface of the specimen for a period of 72 hours. The sample is then split open and the maximum depth of water penetration is measured. The depth of penetration can be difficult to interpret and its measurement is probably accurate only to around ± 5mm. High variability has also been reported even within a single specimen and for different specimens made from the same load of concrete. Laboratories are inconsistent with respect to the choice of the surface to be tested, preparation of the surface and age at test. Very low penetration limits are often specified and the shor tcomings with the test can make them very difficult to apply.
F1.3
BS 1881 Part 122, Water absorption
This test is carried out on 75mm cores at an age around 28 days. The cores are dried for a period of 3 days at 105°C and then immersed in water. Absorption is measured by the increase in weight over a period of 30 minutes. One difficulty with this test is that the specimen may have both a cast face and cut faces and possibly a broken face which will each have different absorption characteristics. These differences could mask the effect of any changes in the cast and cured face which it would be useful to be able to detect. The results are more reliable than other durability tests and the test could be used for routine quality control but is probably not sufficiently sensitive to pick up anything but major variations in concrete quality.
F1.4
BS 1881 Part 208, Initial surface absorption test (ISAT)
The test uses a cap sealed against the concrete surface. The rate of absorption is measured by monitoring the movement of the meniscus along a capillary tube connected to the cap while the water in the cap is maintained at a low constant pressure
42
ASTM C1202/AASHTO T-277, Ability to resist chloride ion penetration (rapid chloride permeability test)
This test measures the passage of electric current through a 50mm thick sample of saturated concrete. The sample is positioned at the centre of a cell with sodium chloride solution on one side and sodium hydroxide solution on the other. An electrical potential of 60V is applied across the concrete sample and the total current passed over a period of 6 hours is measured. The test is slightly less unreliable than the BS EN 12390 Part 8 test in terms of repeatability but has several severe limitations. Most of these limitations are set out in the test s tandard itself which considers the test as only a general guide to durability. The test can show high variability. The test standard advises that differences of 40% can be expected when two samples from the same concrete batch are tested by the same operator. The test is also sensitive to age and to the type of cement used in the concrete. Testing at an age earlier than 28 days is unreliable and it may be better to test at 56 or even 90 days when more consistent results are likely to be obtained. Shi(172) has reported that the test results can be influenced by the type of binder because of alterations in the pore solution chemistry. Hence the criteria given in ASTM C1202 for assessing concrete containing Portland cement do not apply when fly ash, ggbs or silica fume form part of the cement. Shi proposes that a very similar test, namely Nordtest NT Build 492 (171) which uses the same test set-up as ASTM C1202, is more reliable and is not affected by the presence of other cementitious materials. At the end of the test period the sample is split and the depth of chloride penetration is determined by spraying with silver nitrate solution. Unfortunately, there is little or no experience of this test in the Gulf region. Krieg(23) undertook a review of the ASTM C1202 test and concluded that the poor precision associated with the test and the fact that the test solution does not have any influence on the test results effectively invalidates the rapid chloride permeability test as an instrument for quality control. The ASTM C1202 test is convenient in that it is quick to carry out but its fatal drawback is that its results are unreliable. In addition, its results are not in a form which can be used in chloride ingress modelling to predict the rate at which chlorides are likely to enter concrete in service.
Appendix F – Permeability testing
F1.6
Nordtest NT Build 443 accelerated chloride penetration
The principle of this test is that a saturated concrete specimen is exposed to a chloride solution for at least 35 days. After exposure, samples are taken from the exposed face at successive depths by grinding. The individual samples are tested for acid-soluble chloride content. The results are analysed by fitting them to a Fick’s law diffusion equation curve using a least-squares technique. The result is expressed as an effective chloride transport coefficient or a ‘penetration parameter’. This test is unique among those documented here in that it provides a chloride transport parameter which can be used in durability models.
F1.7
Nordtest NT Build 492 chloride migration from non-steady-state migration experiments
The test is a development of the ASTM C1202 rapid chloride permeability test. The concrete sample is exposed to sodium chloride solution on one side and to calcium hydroxide solution on the other. An initial voltage of 30V is applied across the specimen. This voltage is adjusted according to the initial current passed. The test duration varies between 6 and 96 hours depending on the initial current. At the completion of this par t of the test, the specimen is split open and the depth of chloride penetration is determined by spraying the fractured face with silver nitrate. The non-steady-state migration coefficient is calculated from the applied voltage and the average chloride penetration depth.
F2
From these, typical values specified for adverse exposure conditions are set out in Table F2. Note that these are typical values and not recommended values. Property
Test method
Specified limit at 28 days
Water absorption
BS 1881 Part 122
2%
Water penetration
BS EN 12390
10mm
Rapid chloride
ASTM C 1202
2000 coulombs
Table F2 - Typical values for results of durability tests specified for adverse exposure conditions.
F3
Summary
None of the permeability test methods which are commonly used in the region has sufficient precision or reliability for use as a routine quality control test during concrete production. The best approach, if they are to be used, is to use them at mix development stage. Once a mix has been developed which meets the necessary requirements, its performance can be monitored during production by strength tests and by checking the ingredients and their quantities from the batching plant records. The ASTM rapid chloride permeability test could be used as a coarse screen during production, for example as a check that silica fume has been included in a mix, but not as part of the overall compliance requirement.
Specified values
A wide range of values has been observed in concrete specifications from the region as shown in Table F1. These are not to be considered as recommendations. Type
All Ordinary Substructure Silica fume ggbs
All Ordinary Substructure Silica fume ggbs
Minimum
Maximum
Minimum
Maximum
ASTM C 1202 Charge passed (coulombs)
BS EN 12390 Penetration (mm)
700 700 1000 800 800
5 10 10 5 5
3500 3500 1000 2500 1200
50 50 15 10 20
BS 1881 Part 208 10 min flow (ml/m2 /s)
BS 1881 Part 122 Absorption (%)
0.05 0.05 0.15 0.05 0.15
1.0 1.5 1.5 1.0 1.5
0.30 0.30 0.15 0.30 0.15
3.0 3.0 2.0 3.0 2.0
Table F1 - Range of values for results of durability tests found in various specifications.
43