Minimum requirements for durable concrete Carbonation- and chloride-induced corrosion, freeze-thaw attack and chemical attack Edited by D.W. HobbsThe authors would like to thank Jo Bell and Gillian Bond for their advice and assistance in the preparation and layout of this book. The work presented was carried out by the staff of the British Cement Association and the Building Research Establishment. The authors wish to thank the Department of the Environment, Transport and the Regions for providing part-funding of the work under their Partners in Technology Project C1 38/13/9 (CC 719). The views expressed are those of the authors and not necessarily those of the British Cement Association, the Building Research Establishment or the Department of the Environment, Transport and the Regions. Cover photographs: Royal Sovereign Light Tower. Concrete Society Award 1997, Mature Structures Category. Built in 1969. Owner: Trinity Light House Service. Design: Christiani & Neilsen Ltd. in Association with Sir William Halcrow & Partners. Contractor: Christiani & Neilsen Led. Photo: Photoflite, Ashford, Kent, UK. Pedestrian footbridge, Newark, Nottinghamshire. Built: 1915 Description: Single span 275 m, central thickness 150 mm. Design: Mouchel and Partners Ltd. Corrections: Rear cover : Bryan Fig. captions. Chapter 5: MgSO, not Mg,SO, For further information and up-to-date literature ofthis and related subjects contact BCA’s Centre for Concrete Information on (01344) 762676 45.043 Published by First published 1998 British Cement Association ISBN 0 7210 15247 Century House, Telford Avenue Price Group L Crowthorne, Berks RG4S 6YS ‘© British Cement Association 1998 Telephone (01344) 762676 Fax (01344) 761214 All advice or information from the British Cement Association is intended for those who will evaluate the significance and limitations of its contents and take responsibility forts use and application. No liability (including that for negligence) for any loss resulting from such advice or information is accepted. Readers should note that all BCA publications are subject to revision from time to time and should therefore ensure that they are in possession of the latest version.Minimum requirements for durable concrete Carbonation- and chloride-induced corrosion, freeze-thaw attack and chemical attack Edited by D. W. HobbsPREFACE Within the forthcoming European Standard pr EN 206, Conerete- Performance, production and conformity, durability of conerete will rely on prescriptive specification of minimum grade, minimum binder content and maximum water-binder ratio for a series of defined environmental classes. Minimum cover to reinforcement is specified in DD ENV 1992-1-1 (Eurocode 2: Design of concrete structures. Part 1). In the development of pr EN 206, it has not proved possible within the European Committee for Standardization (CEN) to agree common values for the specification parameters to cover the ‘wide range of climates and wide range of cements in use in the EU Member States. The standard will, therefore, almost certainly include indicative values, and it will be left to individual Member States to specify national values where they require them to differ from the indicated values. In the present draft, the indicative values are based on the mean of the range of values currently in use across Europe, and apply only to Portland cements (CEM I cements). No procedure for modifying these values for other cements has yet been agreed. ‘tis thus clear that the prescriptive specification of durable concrete for construction in the UK is entirely dependent upon the establishment of UK national requirements for minimum grade, minimum binder content and maximum water-binder ratio based on values of minimum cover given in DD ENV 1992-1-1. It is not acceptable to base UK national requirements on existing British Standards for two reasons: pr EN 206 uses a new system of exposure classification based on deterioration mechanisms. Existing British Standard exposure classes do not use this system and cannot be related satisfactorily to the new classes. There has been no recent establishment ofa UK consensus on durability requirements as the necessary revision of BS 8110 has not been made because of the higher priority of work on development of the European Standards. There is strong belief that some recommendations in BS 8110 will not provide adequate durability, particularly where resistance to chloride ingress is required, whereas other recommendations may be unnecessarily onerous. ‘The objective of the work presented in this book is to review the UK literature and, where required, the intemational literature to establish the minimum specifications for concrete necessary to achieve durable concrete in the exposure conditions defined by the pr EN 206 for the broader range of binders used in the UK. This necessitated the making of a number of assumptions, and consideration of the minimum requirements presented in this book should include an assessment of these assumptions, to check that they provide an appropriate safety margin for structures in service. The work is being published to stimulate wider informed debate. Due to lack of published data, it was not possible to establish the minimum specification for conerete prepared using all the common binders for each of the pr EN 206 exposure classes ‘The work presented was carried out by the staff of the British Cement Association and the Building Research Establishment and was part-funded by the Department of the Environment, Transport and the Regions under their Partners in Technology programme.CONTENTS ray 12 13 21 22 23 24 25 2.6 2.9 Preface ‘Common cements and concrete exposure classes D. W. Hobbs and J. D. Matthews Common cements 1.1.1 Types of cements (Spooner, 1995) 1.1.2 Constituents of cement (Spooner, 1995) 1.1.3. Strength classes (Corish, 1994) 1.1.4 Sulfate resisting cements to pr ENV 197-X 1.LS Additions (Harrison, 1994) Exposure classes 1.2.1 pr EN 206 (BSI 1997e, d) 1.2.2 BS 8110 (BSI: 1997e) and BS 5328 (BSI: 19971) References Minimum requirements for concrete to resist carbonation-induced corrosion D, W. Hobbs, B. K. Marsh, J. D. Matthews and S. Petit Summary Introduction Methodology Carbonation-induced corrosion in the UK Measurement of carbonation depth Coneretes subject to moderate humidity (XC3) 2.6.1 CEM 1 coneretes (PC) 2.6.2. CEM Vfly ash concretes 2.6.3 CEM Uslag concretes 2.6.4 Minimum specification for concrete subject to exposure class XC3 Coneretes subject to cyclic wetting and drying (XC4) 2.7.1 CEM Iconeretes 2.7.2. CEM Mfly ash concretes 2.73 CEM Vslag concrete 2.7.4. Minimum specific nn for concrete subject to exposure class XC4 Minimum specification for concrete subject to exposure classes XC and xC2 Discussion 2.9.1 Minimum binder content 2.9.2. Influence of curing 2.9.3 Exposure classification 2.9.4 Assumptions made during analysis of available data 2.9.5. Discussion of the assumptions2.10 211 Bu 32 33 34 35 3.6 37 38 39 3.10 3. 3.18 3.19 ‘Minimum requirements for concretes to resist carbonation-induced corrosion References Appendix 2.1: Si fume concretes subject to moderate humidity (XC3) ‘Appendix 2.2: Unpublished BCA and BRE data on the carbonation of CEM I, CEM I/fly ash and CEM Islag coneretes subject to sheltered external exposure Minimum requirements for concrete to resist chloride-induced corrosion D. W. Hobbs and J. D. Matthews Summary Introduction Chloride-induced corrosion in the UK Chloride ingre marine exposure class XS3 Service life and the propagation period: exposure class XS3 Minimum specification for concrete subject to exposure class XS3 3.6.1. Assumptions based on field performance: CEM I coneretes 3.6.2. Minimum specification: CEM I concretes 3.6.3 Assumptions based on field performance: TIVA and IIV/B concretes EM IV/B-S, IV/A-D, 1/B-V, 3.6.4 Minimum specification: CEM I/B-S, I/A-D, I/B-V, IIV/A and I/B coneretes Chloride ingress: marine exposure class XS1 ‘Minimum specification for concrete subject to exposure Class XS1 Chloride ingress into submerged concrete: marine exposure class XS2 Minimum specification for concrete subject to exposure class XS2A Chloride ingress: de-icing salt exposure class XD3 3.11.1 CEM I coneretes 3.11.2 CEM IV/B-V and CEM IIV/A coneretes Service life: exposure class XD3 Minimum specification for concrete subject to external class XD3 exposure 3.13.1 Assumptions based on field performance: CEM I coneretes 3.13.2. Minimum specification: CEM I concretes 3.13.3 Minimum specification: non-CEM I coneretes Chloride ingress: exposure class XDI Chloride ingress: permanently submerged concrete: exposure class XD2A. Assumptions made during analysis of available data. ‘Minimum requirements for concrete to resist chloride-induced corrosion 3.17.1 Chlorides other than sea water 3.17.2 Chlorides from sea water Acknowledgement References iv 4B 43 48 50 35 58 58 604a 42 43 44 45 4.6 47 48 49 4.10 Minimum requirements for concrete to resist freeze-thaw attack D. W. Hobbs, B. K. Marsh and J. D. Matthews Summary Introduction Freeze-thaw attack in the UK Testing 4.4.1 Tests for freeze-thaw resistance of concrete 4.4.2 Tests for scaling resistance of concrete due to freeze-thaw attack in the presence of salt. 4.4.3 Tests on aggregates Air-entrainment and freeze-thaw resistance 4.5.1 Dependence on volume concentration of air 4.5.2 Dependence on ait-void spacing Air-entrained coneretes saturated without salt: XF3 4.6.1 CEM I (and CEM I/SR) concretes 4.6.2. CEM Iifly ash concretes 4.63 CEM Mslag concretes 4.6.4 Portland limestone cement concretes, CEM IV/A-L, B-L 4.6.5 CEM Wsilica fume concretes 4.6.6 Minimum specification for air-entrained concrete subject to exposure class XF3 Air-entrained concretes saturated with salt: XF4 4.7.1 CEM Land CEM 1/SR coneretes: marine exposure 4.7.2 CEM Iislag coneretes: marine exposure 4.73 CEM Ifly ash coneretes: marine exposure 4.7.4 CEM I concretes: de-icing salt exposure 4.7.5 CEM lI/fly ash concretes: de-icing salt exposure 4.7.6 CEM Islag conoretes: de-icing salt exposure 4.7.7 CEM l'silica fume concretes: de-icing salt exposure 4.7.8 Minimum specification for air-entrained concrete subject to exposure class XF4. Non-air-entrained concretes 4.8.1 Performance of high quality concretes 4.8.2 Minimum specification for non-air-entrained conerete subject to exposure classes XF3 and XF4 Minimum requirements for concrete to resist freeze-thaw attack 4.9.1 50 years vs. 100 year required working life 4.9.2 Comparisons with existing specification 1 Relevant exposure classes in British Standards. 2 Comparison of recommendations for minimum concrete quali 4.9.3. Minimum requirements References 1 o1 92 97 97 98 99 99 99 100 103 103 104 104 105 105 106 106 106 108 109 109 110 112 4 4 115 115 118 119 119 120 120 120 121Sa 52 55 5.6 57 58 Minimum requirements for concrete to resist chemical attack D. W. Hobbs and J. D. Matthews Summary Introduction Chemical attack in the UK 5.3.1 Sulfate attack 5.3.2 Acid attack Sulfate attack 5.4.1 General 5.4.2. Criteria for assessment 5.4.2 Portland cements: CEM I and CEM VSR 5.4.3 Minimum specification: CEM I and CI VSR concretes 5.4.4. Portland limestone cements: CEM II/A-L and B-L 5.4.5 Minimum specification: CEM II/A-L and B-L concretes 5.4.6 Cements containing fly ash: CEM IVA-V, B-V, IV/A and IV/B 5.4.7 Minimum specification for cements containing fly ash: CEM II/A-V, B-V, IV/A and IV/B 5.4.8 Cements containing slag: CEM II/A-S, B-S and III/A, B, C 5.4.9 Minimum specification for cements containing slag: CEM I/A-S, B-S and IVA, B, C 5.4.10 Portland silica fume cements: CEM IV/A-D ‘Sea water attack Acid attack 5.6.1 Influence of cement type 5.6.2. Minimum specification for concretes subject to sulfate and acid attack ‘Minimum requirements for concrete to resist sulfate and acid attack 5.7.1 Sulfate attack 5.7.2 Acid attack References Author index Subject index Glossary of cement types: equivalence of British and European cements vi 131 131 132 136 136 137 138 138 138 139 143 144 145 145 150 150 153 153 154 156 156 157 158 158 159 159 163 167 back cover‘Common cements and concrete exposure classes CHAPTER 1 Common cements and concrete exposure classes by D. W. Hobbs, British Cement Association J.D. Matthews, Building Research Establishment 1.1 Common cements L.L.1_ Types of cement (Spooner, 1995) ‘The European prestandard for common cements was published in the UK in 1995 (BS DD ENV 197-1). Since DD ENV 197-1 was published, it has been developed further towards a full European standard, the latest draft of which appeared in March 1997 (pr EN 197-1). Five types of cements are covered, ranging from CEM I (Portland cement) to CEM V (composite cement). Within each of these five types are subdivisions indicating the proportion of Portland clinker, the second main constituent, the standard strength class and rate of strength gain. For example, Portland fly ash cement may be denoted by CEM Il/B-V 32.5N where: CEM IT is the type of cement ~ /B indicates a medium proportion of clinker. A would be higher and C lower; ~ Vis the subtype; indicates the second main constituent, in this case fly ash; - 32.5 is the standard strength class, ~ Nis the sub-class: indicating normal carly strength. R indicates rapid early strength, If left blank, it implies a normal early strength. The cement types in DD ENV 197-1 (BSI, 1995a) and the British Cement Standards are compared in Table 1.1, whilst Table 1.2 gives the cement types covered by British Standards. British Standards do not exactly copy all the chemical requirements of DD ENV 197-1 and an important exception is the requirement for sulfate (as SO,). British Standards retain an effective limit of 3.5 % SO; by mass of cement for all cements.Mini yum Fea sments for durable concrete Table 1.1: Types of cements and their compositions in DD ENV 197-1 Nomenclature | Cement designation | Clinker | Content of other British to DD ENV 197-1 | content Standard (%) cement CEMT Portland cement 95 -100 - BS 12: 1996 CEM I/SR 95 -100 - BS 4027: 1996" CEMIVA-S | Portland-slag cement | 80-94 6-20 BS 146: 1996 CEM II/B-S 65-79 21-35 CEMIVA-D | Portland-silica fume | 90-94 6-10 None cement CEMIVA-P | Portland-pozzolana | 80-94 6-20 ‘None CEMIVB-P | cement 65-79 21-35 ‘None CEM IV/A-Q 80-94 None CEM I/B-Q 65-79 None CEMIVA-V | Portland-fly ash 80-94 6-20 BS 6588: 1996 CEM I'B- cement 65-79 21-35 BS 6588: 1996 CEM IVA-W 80-94 6-20 None CEM II/B-W 65-79 21-35 None CEMIVA-T | Portland-burnt shale | 80-94 6-20 None CEMIVB-T | cement 65-79 21-35 None CEMIVA-L | Portland-limestone | 80-94 6-20 BS 7583: 1996 CEMIVB-L | cement 65-79 21-35 None CEMII/A-M_ | Portland-composite | 80-94 6-20 None CEMIVB-M | cement 65-79 21-35 None CEMIWA — | Blastfumace cement | 35-64 36 - 65 BS 146: 19968 CEM IB 20-34 66 - 80 Nonet CEM TC 5-19 81-95 None’ SEMIVIA | Pozzolanic cement 65-89 11-35 None CEM IViB 45 - 64 BS 6610: 1996 CEM V/A Composite cement 40 - 64 36 - 60 None CEM W/B 20-39 61-80 None * Cement to BS 4027: 1996, Sulfate resisting Portland cement is included here but will eventually be covered specifically in a future part of DD ENV 197-1. 8 In BS 4246: 1996 high slag blastfiurnace cement covers a cement with a blastfumace slag content of 50 - 85%. Note: ASTM Type I to V cements contain no minor additional constituent, ASTM Type I (no C ASTM Type III (C,A $15%) is equivalent to CEM U/R ASTM Type IV (C,A <7%) is equivalent to what may be CEM 1/42.5 MH. ASTM Type V (CA 5%) is equivalent to CEM /SR (ASTM C 150-96), limit) and ASTM Type Il (C,A <8%) are equivalent to CEM IN’Common cements and concrete exposure classes Table 1.2: Cement types in British Standards, BS 12 | Portland cement BS 146 _ | Portland-slag cement BS 146 | Blastfurnace cement BS 4027 | Sulfate-resisting Portland cement BS 4246 | High slag blastfumnace cement BS 6588 | Portland pulverized-fuel ash cements BS 7583. | Portland-limestone cement BS 6610_| Pozzolanic pulverised-fuel ash cement BS $224 _| Masonry cement 1.1.2 Constituents of cement (Spooner, 1995) The possible constituents of factory-produced cements, corresponding to the various cement types listed in Tables 1.1 and 1.2, are as follows: Portland cement clinker Calcium sulfate as a set-regulator Other main constituents (eg. fly ash, slag, limestone) Minor additional constituents (eg. fly ash, slag or limestone [to optimise properties]) Additives (eg. grinding aids, air-entraining improvers, pigments). ‘The code letters used in the DD ENV 197-1 to indicate the type of second main constituent are - blastfurnace slag - silica fume = natural pozzolana + industrial pozzolana (the term ‘artificial’ is used in pr EN 197-1, BSI, 1997) siliceous fly ash (ie. UK fly ash) = calcareous fly ash (ie. high-lime fly ash) ~ burnt shale ~ limestone ~ a composite of two or more of the above gt4e400an€ | >600and | >14a0 and | >3000 and <6000 400, | <600 | 12 and >3tand | >3.8and erecta) | <2 | ats 267 | <67 Vacidextract | 22000and «3000 | >2000and 12000 | >12000 and <24000 mars) pilof groundwater orwater | <65and25.5 40 and £50 >100 water ‘NB, (ng) in water 215 and <30 230 and <60 360 and 100 Mg (me) [Groundwater <1000 ~]_>1000 Brackish 1000 and 53000 3000 groundwater 1.2.2 BS 8110 (BSI, 1997e) and BS 5328 (BSI, 19971) ‘The current UK exposure conditions described in BS $328 (1997f) are reproduced in Table 1.7. Here there are six exposure classes, one of which is when the exposure is abrasive. It could be argued that broadly each of the BS 5328 exposures combines several of the BS EN 206 exposure classes. However, it is not possible to give precise equivalence. The classification of the exposure conditions for sulfate and acid attack taken from BS 5328: 1997 are given in Tables 1.8 and 1.9. Here, there is similarity in the sulfate classes between the modified pr EN 206 chemical exposure classes and BS 5328: 1997 sulfate exposure classes, but for acid attack, the minimum pH in Table 1.6 is 4.0 (BSI, 1997c) as compared with 2.5 given in Table 1.9 (BS 5328: 1997).‘Common cements and concrete exposure classes Table 1.7: Classification of exposure conditions in BS 8110: 1997 and BS 5328: 1997. Environment | Exposure conditions Mild Concrete surfaces protected against weather or aggressive conditions Moderate Exposed concrete surfaces but sheltered from severe rain or freezing whilst wet Concrete surfaces continuously under non-aggressive water Conerete in contact with non-aggressive water Concrete in contact with non-aggressive soil (see class 1 of Table 7a) Concrete subject to condensation Severe Concrete surfaces exposed to severe rain, alternate welting and drying or occasional freezing or severe condensation Very severe Concrete surfaces occasionally exposed to sea water spray or de- salts (directly or indirectly) Concrete surfaces exposed to corrosive fumes or severe freezing conditions whilst wet Most severe Concrete surfaces frequently exposed to sea-water spray or de-icing salts (directly or indirectly) Concrete in sea water tidal zone down to Im below loviest low water Abrasive Concrete surfaces exposed to abrasive action eg. machinery, metal-tyred vehicles or water carrying solids ‘Table 1.8: Sulfate exposure classes (BS 5328: 1997). Sulfate Exposure conditions - Concentration of sulfate and magnesium class In groundwater In soil or fill Byacid extraction | By 2:1 water/soil extract SO, (e/) | Mg (eM) SO, (%) SO. @/) Mg (e/) 1 <0. - <0.24 <12 - a o4to 14 . 1.21023 - 7 uote = Classify on the basis, ee - 4a 3.11060 | <0 ofa 2:1 water/soil 3.81067 «12 extract 4B 3.11060 | >10 3.81067 >12 SA > 6.0 < 10 >67 12 5B >6.0 >10 >67 P12Minimum requirements for durable concrete Table 1.9: Modification to Table 1.8 for concrete exposed to attack from acids in natural ground (BS 5328: 1997). pat Mobility of water | Change in classification with respect to mi content and maximum free water/cement ratio for the cement group recommended on the basis of sulfate class in Table 18 551036 | Static No change Mobile Raise by one sulfate class 3.51025 | Static Raise by one sulfate class Mobile Raise by one sulfate class + Determined by the method given in clause 9 of BS 1377: Part 3, 1990 1.3 References BRITISH STANDARDS INSTITUTION (1995a). Cement-Composition, specifications and conformity criteria. Part 1. Common Cements, DD ENV 197-1: 1995. BRITISH STANDARDS INSTITUTION (1995b). Fly ash for concrete - Definitions, requirements and quality control. BS EN 450: 1995. BRITISH STANDARDS INSTITUTION (19974). Cement - Part 1: Composition, specifications and conformity criteria of common cements. Pr EN 197-1. Document 97/103566, Committee B/516, BRITISH STANDARDS INSTITUTION (19976). Suffate-resisting cements. Pr ENV 197-X. BSI Document 97/103303, Committee B/S16/6. BRITISH STANDARDS INSTITUTION (1997c). Concrete - performance, production and conformity. Pr EN 206, Draft for Public Comment, Document 97/104685, Committee B/517. BRITISH STANDARDS INSTITUTION (19974). Complementary UK concrete standard to pr EN 206, BS XXX: 1998, 12* draft, June, 1997. Committee B/517, Document accompanying BSI 1997a. BRITISH STANDARDS INSTITUTION (1997¢). Structural use of concrete. Part 1. Code of Practice for design and construction. BS 8110: 1997: Part 1 BRITISH STANDARDS INSTITUTION (19979. Concrete Part I. Guide to specifying concrete. BS 5328: 1997: Part 1 CORISH, A. T. (1994). Common cement types and classes (ENV 197-1: 1992). Euro-Cements. Impact of ENV 197 on concrete construction (Editors: R. K, Dhir and M. Roderick Jones). E. & F. N. Spon, pp 11-21. HARRISON, T. A. (1994). Overview of British and European standards for cement and additions. Euro- Cements. Impact of ENV 197 on concrete construction (ditors: R. K. Dhir and M. Roderick Jones). E. &F.N. Spon, pp 1-10. SPOONER, D. C, (1995). The selection of Portland cements to British Standards and to European prestandard ENV 197-1. The Structural Engineer, Vol. 73, No. 20, pp 17-19. 10‘Carbonation-induced corrosion CHAPTER 2 Minimum requirements for concrete to resist carbonation-induced corrosion of reinforcement by D. W. Hobbs, British Cement Association B. K. Marsh, Building Research Establishment (now with Ove Arup and Partners) J.D, Matthews, Building Research Establishment S. Petit, formerly of the British Cement Association 21 Summary ‘The draft European standard pr EN 206 (BSI, 1997a) classifies coneretes subject to a risk of carbonation-induced corrosion according to four exposure classes: xcl xc2 xC3 xc4 Dry Wet Moderate humidity Cyclic rarely dry wet and dry The object of the work described in this chapter is to propose the minimum concrete qualities for these exposure classes which are necessary for design working lives of 50 and 100 years when the minimum cover to reinforcement is 30 mm. This is the minimum cover given in DD ENV 1992-1-1 (BSI, 1992) for exposure classes judged as encompassing similar conditions to XC1 to XC4. It is assumed that the cover concrete is well compacted. The cements considered are CEM I (Portland cement) and binders containing pulverized fuel ash, slag and silica fume. The minimum concrete qualities proposed for exposure classes XC3 and XC4 are applicable to the UK. In the case of CEM I and CEM I/pfa coneretes, they are based primarily on UK studies of the performance of concretes exposed externally after a 24-hour cure and for CEM Uslag concretes, after a three-day cure. For exposure classes XCI and XC2, it is proposed that a nominal minimum quality of conerete will be suitable for both. i‘Minimum requirements for durable concrete 22 Introduction When carbon dioxide from the atmosphere diffuses into concrete, it combines with water, forming carbonic acid, which then reacts with alkali hydroxides forming carbonates. In the presence of free water, calcium carbonate is deposited in the pores of the concrete at the depth at which carbonation is occurring. As a consequence of carbonation, the pH of the pore fluid drops from a value greater than 12.6, in the uncarbonated region, to a value of about & in the region of complete carbonation. If this reduction in alkalinity occurs close to the steel, it can cause its depassivation. In the presence of moisture and oxygen this can lead to corrosion followed by corrosion-induced cracks parallel to reinforcement, followed by spalling. Such deterioration can occur because the corrosion product which forms can be up to 2 to 4 times the volume of steel before it oxidized. A model of the effect of carbonation-induced corrosion upon service life is shown in Figure 2.1 (Tuutti, 1982). ‘There is an initiation period, which is the period taken for the carbonation front to reach or approach the steel. This is followed by the propagation period which is the period between initiation of corrosion and cracking Controlled by : Controlled by | difasion rmostre state cover , avaliily ot cathode | we OF fon concentaion | caring temperatire temperature Depth of corrrasion | | binder type | | | I | | CO, ingress Initiation Propagation Lifetime or time before repair Figure 2.1: Corrosion model, After Tuutti (1982) In practice, carbonation-induced corrosion is generally regarded as a minor problem compared with chloride-induced corrosion (Parrott, 1987). However, the bulk of structural conerete is used in residential, industrial and office buildings, rather than engineering structures. Consequently only a small proportion of structures are at risk due to chloride-induced corrosion. Thus the potential service life of most reinforced concrete is governed by the rate of carbonation, the cover to reinforcement and the rate of reinforcement corrosion. Most research on carbonation has been. restricted to plain concrete. Relatively few studies have been made of carbonation-induced corrosion (Parrott, 1994; Dhir et al, 1992; Schiessl et al, 1994) ‘The draft European standard pr EN 206 classifies concretes subject to a risk of carbonation- induced corrosion according to four exposure classes (BSI, 1997) as follows:Carbonation-induced corrosion xc xC2 x3 xc4 Dry Wet ‘Moderate humidity Cyclic rarely dry wet and dry Examples of concretes which fall into cach of these exposure classes are given in pr EN 206: 1997 as: XC1 Concrete inside buildings with low air humidity. xc2 Parts of water-retaining structures. Many foundations. xC3 Conerete inside buildings with moderate or high humidity, Extemal conerete sheltered from rain. xc4 Surfaces subject to water contact, not within exposure class XC2. ‘The informative examples applicable in the UK, which appear in the complementary UK concrete standard to pr EN 206 (BSI, 1997b) are as follows: xc! Reinforced and prestressed concrete surfaces inside structures, except areas of structures with high humidity. x2 Reinforced and prestressed concrete completely buried in non-aggressive soil, Reinforced and prestressed concrete surfaces permanently submerged. XC3/XC4 External reinforced and prestressed concrete surfaces including those sheltered from direct Reinforced and prestressed concrete surfaces inside structures with high humidity (eg. bathrooms, kitchens). Reinforced and prestressed concrete surfaces exposed to alternate wetting and drying. The object of the work described in this chapter is to propose the minimum concrete qualities for these exposure classes which are necessary for design working lives of 50 and 100 years ie. Classes 3 and 4, respectively of ENV 1991-1: 1994 (European Committee for Standardization, [1994}) when the actual minimum cover to reinforcement is 30 mm. This is the minimum cover given in DD ENV 1992-1-1: 1992 (BSI, 1992) for exposure classes judged as encompassing. similar conditions to XC1 to XC4, It is assumed that the cover concrete is well compacted. ‘The cements considered are CEM I (Portland cement) and binders containing pulverized fuel ash and ground granulated blastfurnace slag. Binders containing silica fume are briefly dealt with in ‘Appendix 2.1. Throughout the remainder of this chapter, pulverized fuel ash is referred to as fly ash or pfa, and ground granulated blastfurnace slag is referred to as slag, The minimum concrete qualities proposed are applicable to the UK and are based primarily on UK studies of the performance of concretes exposed externally after a 24-hour cute or, in the case of concretes containing more than 35% slag by mass of cement, after a three-day cure. Such differences in curing compensate for slag conerete’s slower strength development (BSI, 1997c). The minimum. concrete qualities proposed are not necessarily applicable to the temperatures and humidities of exposure experienced by conerete structures elsewhere in Europe. 13Minimum ments for durable concrete req 2.3 Methodology In this chapter, the limiting mix parameters for concrete quality to provide adequate resistance to carbonation-induced corrosion are based on experimental results reported in the published literature. UK literature was searched for carbonation depth data derived from: experiments using realistic curing and exposure conditions. Where necessary, UK results have been ‘supplemented by results from investigations performed overseas. The text indicates where this has been done. Results are generally available only for carbonation depths up to a few years of age. It has thus been necessary to use established relationships between carbonation depth and age to predict carbonation depths at later ages appropriate to the required design working life given in ENV 191-1: 1994 (European Committee for Standardization, 1994) Damage due to carbonation-induced corrosion depends not only upon the carbonation front approaching or reaching the reinforcement butalso upon the development of sufficient expansion due to corrosion to cause distuption of the concrete cover. Consideration of appropriate concrete qualities has thus necessarily considered rates of both carbonation and corrosion. ‘Various assumptions have had to be made about the carbonation and corrosion rates in reinforced concrete over long term exposure. These assumptions are detailed in the text, but have generally been conservative to allow for differences in behaviour between test specimens and concrete in structures which will generally be more variable in its properties, Results have been analysed to find individual relationships between carbonation depthand water- binder ratio and between carbonation depth and compressive strength, Insufficient relevant information is available to establish relationships between carbonation depth and binder content. Results from coneretes containing fly ash and slag have each been treated separately from those containing CEM I only. It has been assumed that concrete production and placement has followed reasonable good practice and the properties of concrete in the structure are within normal limits of variabil Nevertheless, it has been felt necessary to assume only minimal levels of curing as itis generally agreed that curing is not given adequate attention for much in-situ conerete construction. Determination of required concrete qualities has been based on minimum values of cover to reinforcement as given in DD ENV 1992-1-1 (BSI, 1992). The implications of this on specification of nominal cover are discussed in the sub-Section 2.9.5 entitled ‘Discussion of the assumptions’. 24 Carbonation-induced corrosion in the UK In 1987 Parrott reviewed the reported cases of reinforced concrete members in the UK which had deteriorated due to carbonation-induced corrosion. These included: Bridges where corrosion had occurred on soffits, parapets and on stirrups of beams. Prefabricated houses built in the 1940s and 1950s. DA number of industrialized and system-built dwellings dating from the 1960s and 1970s. Examples of deterioration due to carbonation-induced corrosion are shown in Figures 2.2 and 2.3. 14‘Carbonation-induced corrosion Figure 2.3: An example of deterioration in a building caused by carbonation-induced corrosion. 2.5 Measurement of carbonation depth ‘The depth of the carbonation front in concrete can be measured in a variety of ways: by X-ray diffraction, infra-red spectroscopy, microscopy, thermogravimetric analysis, chemical analysis or the phenolphthalein test. The phenolphthalein testis the simplest and most popular test, and is carried out by spraying a freshly split surface of the concrete with a mixture of phenolphthalein, water and alcohol. This pH indicator changes from colourless to magenta as the pH rises from 8.3 to 10.0. Ona broken 15Minimum requirements for durable concrete concrete surface the change in colour becomes readily visible at a pH greater than about 9.0 to 9.5. The test indicates the depth to which the calcium hydroxide has been depleted. The phenolphthalein indicator test is carried out on a split surface of concrete (rather than on a cut surface or on drillings) as this minimises the movement of calcium hydroxide into the carbonated zone. In most of the papers referred to in this chapter, the depth of carbonation was measured using the phenolphthalein test. 26 Concretes subject to moderate humidity (XC3) For concretes exposed to a moderate humidity, the highest risk of carbonation-induced corrosion exists forextemal exposed concrete sheltered from rain. This section deals with the reported data for this particular exposure regime. 2.6.1 CEM I coneretes (PC) For coneretes subject to sheltered external exposure in the UK, the average long-term relative humidity to which the concretes are exposed is about 80 to 85% (Parrott, 1995), At a humidity of 80%, Parrott has deduced that the depth of carbonation, d,, is related to concrete age, t, by the expression deat! = 2 (QA) Assuming that corrosion commences when d, equals the minimum reinforcement depth, Parrott also deduced that when concrete is subject to long-term exposure at 80% RH, cracking of the concrete, initiated by carbonation-induced corrosion, will commence approximately 20 years after the carbonation front reaches the steel. Data reported by Brown (1987) broadly supports this deduction. Brown examined seven bridges which ranged in age from 51 to 55 years. In nine of the elements examined, carbonation-induced cracking had occurred. The mean ratio of carbonation depth to cover was 1.24 which, from the expression (hy = 1.24 (2.2) gives an upper bound for the corrosion period of 22 years (taking t, to be 53 years gives t, the age at which the carbonation front reaches the steel as 31 years). In the work described in the current section of this chapter, a more conservative approach than deduced by Parrott is adopted, namely that cracking will be induced 15 years after the carbonation front reaches the steel. Thus for design lives of 50 and 100 years, the carbonation depth should not exceed 30 mm at 35 and 85 years respectively. Several investigators in the UK have measured the depth of carbonation into laboratory produced CEM I concretes subject to sheltered external exposure (Hobbs, 1988; Thomas and Matthews, 1992; Parrott, 1990; Barker and Matthews, 1994; Matthews, 1994). Additional unpublished data obtained by BCA and BRE are summarized in Appendix 2.2 (Hobbs, 1995; Matthews, 1995; Brown and Beeby, 1992). The carbonation depths obtained are shown plotted against free water-cement ratio in Figure 2.4, together with some results obtained on coneretes exposed to an RH of 80% (Byfors, 1985). The carbonation depths plotted in Figure 2.4 have been normalized to ages of 35 and 85 years using equation 2. . For concretes subject to a one-day cure, examination of Figure 2.4 indicates that, on the basis of the experimental data, maximum free ‘water-cement ratios of 0.61 and 0.45 should ensure that the depth of carbonation is unlikely to 16Carbonation-induced corrosion 20 © tdcure © 3dcure 4 7d cure, Byfors (1985) Normalized carbonation depth at 35 years (mm) Normalized carbonation depth at 85 years (mm) ee) Tw wic ratio Figure 2.4: Dependence of carbonation depth upon free w-c ratio. CEM I coneretes. Sheltered external exposure. UK data. exceed 30 mm until after ages of 35 and 85 years respectively. Data Jeading to a similar conclusion has been obtained by overseas investigators (Figure 2.5, Byfors, 1985; Wierig, 1984; Hakkinen, 1993; Fagerlund, 1988; Meyer, 1968). In Figure 2.6 the measured depths of carbonation into sections of seven UK bridges, are shown plotted against deduced free water-cement ratio (Brown, 1987). The ages of the bridges ranged from 51 to 55 years and the elements were subject to class XC3 or XC4 exposure. The free ‘water-cement ratio was deduced from the capillary porosity and cement content and, according to Brown (1987), has an accuracy of about * 0.1. Also shown plotted in Figure 2.6 is the upper bound curve taken from Figure 2.4, normalized to an age of $3 years. The normalized curve is a reasonable upper bound to the carbonation depths measured by Brown. In Figure 2.7 the depth of carbonation measured by UK investigators is shown plotted against the 28-day cube compressive strength (Hobbs, 1988; Thomas and Matthews, 1992; Parrott, 1990; Barker and Matthews, 1994; Matthews, 1994, 1997; Hobbs, 1997; Brown and Beeby, 1992). For concretes cured for one day and ignoring one outlier, minimum concrete strengths of 36 and 52 N/mm? should ensure that, on the basis of the experimental data, the depth of carbonation is, unlikely to exceed 30 mm until after ages of 35 and 85 years respectively. If the outlier is included, then these strengths become 45 and 54 N/mm’, 7‘Minimum requirements for durable concrete 70) 60| 50) 40] 30] 20| Normalized carbonation depth at 35 years (mm) 10] Wierig, Germany, 7d cure, 2,8, 16 year Hakkinen, Finland, 10d cure, 2, 4 year Byfors, Sweden, 7é cure, 2.5 year Fagertund, Sweden, 1 and 7d cure, 1 yeer Meyer, Germany, 1d cure, 3 year 04 O85 O08 OF 08 09 we ratio, Normalized carbonation depth at 85 years (mm) Figure 2.5: Dependence of carbonation depth upon w-c ratio, CEM I concretes. external exposure. Non-UK data, Sheltered Carbonation depth (mm) 60 wic ratio Figure 2.6: Dependence of carbonation depth into sections of seven UK bridges upon deduced water-cement ratio, Ages 51 to 55 years, CEM I concretes, 18Carbonation-induced corrosion Normalized carbonation depth at 35 years (mm) Normalized carbonation depth at 85 years (mm) lo 20 30 40 50 60 70 28d compressive strength (Nimm?) Figure 2.7: Dependence of carbonation depth upon 28d compressive strength. CEM I concretes. Sheltered external exposure, UK data. 2.6.2. CEM Mfly ash concretes Carbonation data obtained by UK investigators on CEM I/fly ash coneretes subject to sheltered extemal exposure are plotted against free water-binder ratio in Figures 2.8, 2.9 and 2.10 (Hobbs, 1988; Thomas and Matthews, 1992; Parrott, 1990; Barker and Matthews, 1994; Matthews, 1994, 1995; Hobbs, 1995; Brown and Beeby, 1992). Figure 2.8 plots results for concretes containing 30 or 35% fly ash, Figure 2.9a for 0, 30 and 50% fly ash and Figure 2.10 for 50% fly ash.. Results obtained by overseas investigators (Byfors, 1985; Fagerlund, 1988) on coneretes containing 30% fly ash are shown in Figure 2.11. As in the previous section, the depths of carbonation which are plotted in Figures 2.8, 2.10 and 2.11 have been normalized to ages of 35 and 85 years, A comparison of Figures 2.8, 2.9a and 2.10 with Figures 2.4 and 2.5 shows that, based on free water-binder ratio, the resistance to carbonation of CEM I and CEM I/fly ash concretes is, different. Their differing effects can be estimated by expressing the depth of carbonation at a given age in terms of a water-effective-cement ratio, as foliows: d= b (wife + kp]) — wee (2.3) where p is the mass of fly ash and k, the cementing efficiency factor, is the mass of CEM I cement which gives a similar resistance to carbonation as a unit mass of fly ash. In the case of data published by Thomas and Matthews (1992), the depths of carbonation are similar in CEM Tconeretes and CEM I/fly ash coneretes when (wleesray = [Well + OP Mpc, . QA) 19Minimum requirements for durable concrete Normalized carbonation depth at 35 years (mm) ‘Normalized carbonation depth at 85 years (mm) OSes 0-40-6008 ‘wb ratio Figure 2.8: Dependence of carbonation depth upon free w-b ratio. 30 - 35% fly ash. Sheltered external exposure. UK data 25 08 0% pla Cm 30% pfa 0 50% pfa ome td moist cure (0090 3d moist cure = 20 € = 3 31s § 2 2d y 8 fe 5 % S 6 “5 (a) (o) @ 02 O4 06 O08 02 o4 O6 08 wib wile + 0-1p) Figure 2.9 a & b: Dependence of carbonation depth on free w/b and free wi(c + 0.1p) for coneretes subject to sheltered extemal exposure (after Thomas and Matthews, 1992). 20Carbonation-induced corrosion Normalized carbonation depth at 36 years (mm) Normalized carbonation depth at 85 years (mm) © td cure, 33 yr © 3dcure, 3.3 yr ‘A 1dcure, 2 yr 4 3d cure, 2 yr 0.30 0.35 0.40 0.45 050 (0.55 wib Figure 2.10: Dependence of carbonation depth upon free w-b ratio. 50% fly ash. Sheltered external exposure. UK data. ie, when the k value is 0.1 (Figure 2.96). Carbonation data obtained on coneretes containing 30% fly ash subject to laboratory exposure (Hobbs, 1988; Thomas and Matthews, 1992) suggest a higher value for k (see Figure 2.12a & b). Examination of Figure 2.8 shows that, for a concrete containing 30% fly ash subject to a one-day cure, maximum free water-binder ratios of 0.46 and 0.40 should ensure that the depth of carbonation is unlikely to execed 30 mm until after 35 and 85 years respectively. Analysis of data from overseas investigations, as shown in Figure 2.11, leads to similar maximum values for water-binder ratios, For concretes containing 50% fly ash, the maximum free water-binder ratios are 0.35 and 0.28 respectively (see Figure 2.10). In Figures 2.13 and 2.14 the carbonation data obtained by UK investigators, normalized to ages of 35 and 85 years, is shown plotted against 28-day compressive strength for concretes containing 30 to 35% and 50% fly ash respectively (Hobbs 1988; Thomas and Matthews, 1992; Parrott, 1990; Barker and Matthews, 1994; Matthews, 1994, 1995; Hobbs, 1995; Brown and Beeby, 1992). Examination of these figures shows that, to ensure that the carbonation depth does not exceed 30 mm at 35 and 85 years, minimum 28-day compressive strengths of 44 and 54 N/mm? respectively are required for concretes containing 30% fly ash, subject to a one-day cure, and 49 and 58 N/mm? respectively for concretes containing 50% fly ash. 2Minimum requirements for durable concrete Byfors (1995) Sweden, 7¢ cure, 2.5 yr Fagerlund (1988), Sweden, 1 and 7d.cure, 4 yr 60 30 Normalized carbonation depth at 35 years (mm) 8 & 10 Normalized carbonation depth at 86 years (mm) 03 04 05 0.6 07 08 09 70? wib ratio Figure 2.11: Dependence of carbonation depth upon w-b ratio, 30% fly ash, Sheltered external exposure. Non-UK data. a © 0% pfa OW 30% pfa 04 50% pla @M@ td moist cure ©0O 3d moist cure + é a 8 2 year carbonation depth (mm) 0 02 o4 06 08 04 06 08 w/b witc + 0-8p) Figure 2.12 a & b: Dependence of depth of carbonation on free w-b and free w/(c + 0.3p) for concretes stored at 20°C, 65% RH. (After Thomas and Matthews, 1992). 22Carbonation-induced corrosion 50 © tdcure © 3dcure 40) 30 Normalized carbonation depth at 35 years (mm) Normalized carbonation depth at 85 years (rm) 0 3 a0 30 80 70 30 284. compressive strength (Nim?) Figure 2.13: Dependence of carbonation depth upon 28-day compressive strength, 30-35% fly ash. Sheltered extemal exposure. UK data, 60] = 20} ooh om pee 19] Normatized carbonation depth at 36 years (mm) Normalized carbonation depth at 85 years (mm) o lo 20 ED} 40 30 60! 286 compressive strength (N/mm?) Figure 2.14: Dependence of carbonation depth upon 28d compressive strength, 50% fly ash Sheltered external exposure. UK data. (After Thomas and Matthews, 1992; Hobbs, 1995, Appendix 2.2). 23Minimum requirements for durable concrete 2.6.3. CEM Uslag concretes Carbonation data obtained by UK investigators on CEM I/slag coneretes with slag contents of 50 and 70% subject to sheltered external exposure are plotted against free water-binder ratio in Figure 2.15 (Parrott, 1990; Barker and Matthews, 1994; Hobbs, 1995; Matthews, 1995; Osborne, 1994), Also included in Figure 2.15 are data obtained from field structural elements (Osbome. 1994), Results obtained by overseas investigators are shown plotted against water-binder ratio in Figure 2.16 (Wierig, 1984; Hakkinen, 1993; Meyer, 1968; Kleinschmidt, 1965). As in the previous sections, the depths of carbonation are normalized to ages of 35 and 85 years. In Figure 2.17 the depth of carbonation determined by UK investigators is plotted against 28-day cube compressive strength together with data obtained on companion CEM I coneretes (Barker and Matthews, 1994; Hobbs, 1995; Matthews, 1995; Brown and Beeby, 1992), A comparison of Figures 2.15 and 2.16 with Figures 2.4 and 2.5, and of Figure 2.17 with Figure 2.7, shows that, based on both water-binder ratio or 28-day compressive strength, the resistance to carbonation of CEM I and CEM Uislag concretesis different, irrespective of whether the CEM Vslag coneretes are cured for one or three days. However, the data obtained on CEM Wslag concretes by UK investigators is less comprehensive than that obtained on CEM I and CEM I/pfa concretes, Examination of Figure 2.15 for a concrete containing 70% slag by mass of binder, subject to a three-day cure, indicates that, on the basis of the experimental data, maximum free water-binder ratios of 0.45 and 0.32 may ensure that the depth of carbonation is unlikely to execed 30 mm until after ages of 35 and 85 years respectively. This compares closely with water- binder ratios of 0.45 and 0.34 deduced from the overseas data shown in Figure 2.16. However, the validity of these values may be questionable firstly, because of the narrow range of free 60; © td cure © 3d cure ‘4 Structural elements 8 a & 8 3 Normalized carbonation depth at 85 years (mm) Normalized carbonation depth at 35 years (mm) 8 VE ey wb ratio Figure 2.15: Dependence of carbonation depth upon w-b ratio. 50 and 70% slag. Sheltered external exposure, UK data. 24Carbonation-induced corrosion © Kleinschmidt, Germany, 50% slag, 17 years © Wierig, Germany, 50% slag, 7d cure ‘© Hakkinen, Finland, 70% slag, 10d cure, 2, 4, years. ‘© Wierig, Germany, 78% slag, 7d cure, 2,8, 16 years © Meyer, Germany, 50% slag, td cure, 3 years 60; x 2 = 2 8 g & 8 Normalized carbonation depth at 35 years (mm) Normalized carbonation depth at 85 years (mm) ol 035040580: 0n 00 wib ratio Figure 2.16: Dependence of carbonation depth upon w-b ratio. 50 and 70% slag. Sheltered external exposure. Non-UK data. 60 A 70% td cure . 70% 3d cure 50% 1d cure 50] 1D 50% 3d cure © 40% 1d cure © 40% 3d cure © 0% 1d cure Normalized carbonation depth at 35 years (mm) Normalized carbonation depth at 85 years (mm) 20 30 40 60 60 7 28d compressive strength (N/mm?) Figure 2.17: Dependence of carbonation depth upon 28d compressive strength. CEM V/slag coneretes. Sheltered external exposure. UK data. 25Minimum requirements for durable concrete. wwater-binder ratios covered in the UK work and secondly, because in the overseas work many of the coneretes tested were cured for seven or ten days prior to exposure. Estimates of the possible maximum free water-binder ratios necessary to ensure that the depth of carbonation does not exceed 30 mm until after ages of 35 and 85 years for concretes containing less (or more) than 70% slag can be made using the k-factor approach. From the limited data shown in Figure 2.15, it has been deduced that coneretes, cured for three days, with 70% slag and water-binder ratios of 0.45 and 0.32 give equivalent performance to CEM coneretes, cured for one day, with free water-cement ratios of 0.61 and 0.45 respectively. ‘Comparing these slag concretes and CEM I concretes indicates a k-factor in the range 0.59 to 0.62%. Assuming a k-factor of 0.60 and a slag content of 40% gives water-binder ratios of 0.51 and 0.38 as being equivalent to water-CEM I cement ratios of 0.61 and 0.45 respectively. Inthe case of compressive strength, the data is insuflicient to enable estimates to be made of the minimum 28-day compressive strengths necessary to ensure that the carbonation depth does not exceed 30 mmat 35 or 85 years. At equal 28-day compressive strengths, the limited data indicate that the carbonation depth in CEM V/slag coneretes, subject toa one- or three-day cure, is greater than the carbonation depth in CEM I concretes subject to a one-day cure. 2.6.4 Minimum specification for concrete subject to exposure class XC3 ‘Table 2.1 gives the minimum qualities for concretes subject to class XC3 exposure necessary for design lives of 50 and 100 years, when the minimum cover to reinforcement is 30 mm. For CEM I and CEM pfa coneretes, the minimum qualities assume a one-day cure and for the CEM Vslag coneretes, a three-day cure. ‘Table 2.1: Minimum concrete qualities from experimental data, Exposure class XC3. Conerete mix details | Design ‘Moderate humidity - XC3 life (years | CEM |CEMV30| CEMI/50 | CEM U0 | CEM 170 (fly |(%)flyash| (%) slag | (%) slag ash Minimum compressive] 50. | 36(45)* | 44 49 45 strength (N/mm") Maximum w-b 061 0.46 035 0.51 0.43 Minimum compressive] 100 | 52(54)* | 54 38 (608 strength (Nimm*) Maximum w-b 0.45 0.40 0.28 038 032 * When outlier included. * Approximate 28d compressive strength. 2 For a one-day cure, the k-factor for slag is approximately 0.4 26Carbonation-induced corrosion 27 Coneretes subject to cyclic wetting and drying (XC4) 2.7.1 CEM I concretes Due to the higher moisture content of concrete subject to unsheltered extemal exposure in the UK, the rate of carbonation will be slower than for similar concrete subject to sheltered external exposure. However, if the carbonation front reaches the steel, and corrosion is consequently initiated, it will occur more rapidly. The corrosion rate, CR, the steel reinforcement corrosion period until cracking, t,, and the power n in the generalised form of equation (2.1), i. qdat 7 (2.5) depend upon relative humidity. The estimated values have been determined by Parrott (1995) and are given in Table 2.2, again assuming that corrosion commences when the carbonation depth, 4,, equals the minimum reinforcement depth. Table 2.2: Values of CR (jmv/yr), t, (yr) and the power n. RH (%) 80 90 bal 98 100 cR 5 10 20 50 10 4 20 10 5 2 10 n 0.415 0.317 0.256 0.216 0.187 For concretes subject to unsheltered extemal exposure, the rates of carbonation and subsequent corrosion rates are likely to be greatly influenced by the local microclimate, particularly with respect to moisture condition, and for the purposes of design, onerous values for n and t, should be selected. In the work described in this section of this chapter, itis assumed that t, is zero. Thus for a design life of 50 years, the carbonation depth should not exceed 30 mm at 50 years. To normalize the reported carbonation data to a common age of 50 or 100 years, an onerous upper bound value for n of 0.4 has been chosen (see Table 2.2). A number of investigators in the UK have measured the depth of carbonation into externally exposed CEM I concretes unprotected from rain (Osborne, 1989; Matthews, 1984, 1995; Baweja et al, 1987; Dunster 1991/92; Thomas and Matthews, 1991; Thomas et al, 1990; Bamforth, 1997). The carbonation depths, normalized to ages of 50 and 100 years are plotted against free ‘water-cement ratio in Figure 2.18. A comparison of the data in this figure with that in Figure 2.4 indicates substantially lower carbonation depths for exposed concretes than for those sheltered from the elements. Similar results have been obtained by overseas investigators (compare Figure 2.19 with Figures 2.4 and 2.5), (Wierig, 1984; Fagerlund, 1988; Nagataki et al, 1986; Yoda, 1982; Kokubu and Nagataki, 1989; Frey, 1993; Nishi, 1962; Hamada, 1969). Itfollows that XC4 exposure is less severe than XC3 exposure, despite the onerous assumptions that t, was zero and the power n was 0.4. Examination of Figure 2.18 does not show a clear dependence of the carbonation depth upon free water-binder ratio, Even at a high free water-cement ratio of 0.9, it is deduced that the carbonation depth is unlikely to exceed 30 mm at 100 years. However, overseas data do indicate that carbonation depths could exceed 30 mm at 100 years, Examination of Figure 2.19 indicates that on the basis of the overseas experimental data, maximum water- cement ratios of 0.85 and 0.55 should ensure that the depth of carbonation is unlikely to exceed 30 mm until after ages of 50 and 100 years respectively, 2Minimum requirements for durable concrete § 8 8 Normalized carbonation depth at 50 years (mm) 30) ° 20) ° ° 7 o ° 8 ° — se w/e ratio 08 1 8 Normalized carbonation depth at 100 years (mm) 1 8 eo Extemal exposure. UK data Figure 2.18: Dependence of carbonation depth upon w-c ratio. CEM I coneretes, 2 8 2 g rs & sol 20) 10} Normalized carbonation depth at 50 years (mm) Ga 0a oe (oe 07 wie ratio 8 8 Normalized carbonation depth at 100 years Figure 2.19: Dependence of carbonation depth upon w-c ratio. CEM I coneretes. External exposure. Non-UK data. 2.7.2. CEM Wily ash concretes Carbonation data obtained by UK investigators (Bamforth, 1997; Thomas and Matthews, 1992; Baweja ct al, 1987; Thomas et al, 1990; Matthews, 1995) and overseas investigators (Nagataki 28Carbonation-induced corrosion etal, 1986; Fagerlund, 1988; Kokubu and Nagataki, 1989; Baweja et al, 1987) on CEM I/fly ash coneretes subject to external exposure are plotted against water-binder ratio in Figures 2.20 and 2.21 respectively. A comparison of these figures with Figures 2.18 and 2.19 shows that, based on water-binder ratio, the resistance to carbonation of CEM I and CEM I/fly ash coneretes is different. For a concrete containing 30% fly ash by mass of binder, examination of both Figures 2.20 and 2.21 shows that, on the basis of the experimental data, maximum water-binder ratios of 0.60 and 0.50 should ensure a minimum design life of 50 and 100 years respectively. The g 8 Bamforth (1997), 9 years Baweja et al (1987), 10, 26 years, ‘Thomas and Matthews (1992), 2.5 and 10 years ‘Thomas etal (1990), 2.5 years Matthews (1995), 14 years 20 ‘A 50% pfa (Matthews, 1995) 10| ‘Normalized carbonation depth at 100 years (mm) Normalized carbonation depth at 50 years (mm) os a w/b ratio Figure 2.20: Dependence of carbonation depth upon w-b ratio. 24 to 30% fly ash. External exposure. UK data. g g & mn 8 8 3 Normalized carbonation depth at 50 years (mm) Normalized carbonation depth at 100 years (mm) 8 ° O-6 OT O8 og oe 7 ° w/o ratio Figure 2.21: Dependence of carbonation depth upon w-b ratio. 26 to 30% fly ash. External exposure. Non-UK data. 29Mini jum requirements for durable concrete arrows on Figures 2.20 and 2.21 indicate that these water-binder ratios are safe both for UK and overseas exposure conditions. 2.7.3 CEM Uslag coneretes Carbonation data obtained by Osborne (1989) on CEM I/slag concretes subject to external exposure are shown plotted against free water-binder ratio in Figure 2.22a, It should be noted that, the data are very limited. An examination of Figure 2.22a shows that the resistance to carbonation of CEM I and CEM Islag concretes is different. For example, concretes containing, 40 and 70% slag and with free water-binder ratios of approximately 0.60 to 0.70 and 0.40 to 0.45 respectively, give equal carbonation performance to a CEM I concrete with a free water-cement ratio of 0.85. For CEM I concretes with a free water-binder ratio of 0.55, the comparable free water-binder ratios are in the ranges 0.40 to 0.45 and 0.35 to 0.40 respectively. For concretes subject to XC4 exposure, this indicates a k-factor for slag in the range 0.30 to 0.45 (see Figure 2.226). This range of k-factors is surprising, as higher values than those for exposure class XC3 would be expected. 08 o% slag 3. 40% slag oo Tom st = insulated on all sides for 7 days i: E 8.2.0 two opposite faces uninsuleted a for 7 days & [Siructure eiemonis o bo) normalized to 2.5 years : 0% slag, 7, 12.179 | : X sowsiog.5.y.” | 3 * 3 70% slag, Sy #4 = amy 8 > ta) x tb) So 1 1 1 1 : o4 06 os 08 06 08 10 72 wip w/lo+ 0-48) Figure 2.22a & b: Dependence of depth of carbonation on free w-b ratio and w-{c + 0.45s) ratio for CEM Land CEM I’slag concretes stored externally for 2.5 years. 300 mm cubes top face. 2.7.4 Minimum specification for concrete subject to exposure class XC4 From the previous sub-Sections, the maximum free water-binder ratios necessary for design lives of $0 and 100 years, when the minimum cover to reinforcement is 30 mm, have been deduced for concretes subject to exposure class XC4, The maximum free water-binder ratios correspond in some instances to concrete grades which are lower than the minimum specified for reinforced concrete, namely C24/30. Table 2.3 gives the minimum conerete qualities required for design lives of 50 and 100 years, when the concrete is subject to exposure class XC4.‘Carbonation-induced corrosion Table 2.3: Minimum concrete qualities from experimental data. Exposure class XC4. Concrete mix | Desiga Cyelie wet and dry details | life (yrs) cEMI cEM130 | CEMI40 | CEM¥70 (%) fly ash_ | _(%) slag (%) slag Maximum wh | 50 0.85 0.60 0.60t00.70 | 0.40 t00.45 Maximum wie | 100 035 0.50 04010045 | 0.3510 0.40 28 Minimum specification for concrete subject to exposure classes XC1 and XC2 For coneretes subject to exposure class XC1 (dry), carbonation will occur but carbonation- induced corrosion is unlikely. For coneretes subject to exposure class XC2 (rarely dry), little carbonation will occur. Consequently a nominal minimum quality of concrete will be suitable for both of these exposure classes. 29 Discussion 29.1 Mi The binder content of concrete is generally dependent upon the method of mix design employed rather than on a judgement of the minimum amount required to provide a given performance. Little information is available within published literature on the dependence of carbonation rate upon binder content independent of free water-binder ratio. In the absence of any admixtures, the binder content required at a given free water-binder ratio or compressive strength is dependent upon the water demand of the concrete mix to provide the required level of workability. In practice, water-reducing admixtures are often employed to modify the water demand and thus provide the required level of workability at reduced binder contents. The majority of concretes used in the experiments from which the data in this chapter have been obtained, contained no water-reducing admixtures. Binder contents employed may thus have been greater than may actually have been required. The binder contents of the coneretes tested which corresponded most closely to the coneretes given in Table 2.1 are broadly as follows: um binder content CEMI © 280 kg/m? for wie 0.60 370 kg/m’ for wie 0.45, CEM 1/30 % pfa ¢ 350 ke/m? for wie 0.45 © 400 kg/m’ for w/e 0.40 CEM 1/50 % pfa 400 ke/ny for w/e 0.35 CEM 1/50 % slag © 360 kg/m? for wie 0.50 CEM 1/70 % slag © 360 kg/m? for wie 0.50 where ¢ is the cement or binder content (kg/m), Recommendations for minimum binder contents given in current British Standards, such as BS 8110 for the structural use of concrete, are generally high compared with many national standards in other CEN (European Committee for Standardization member states). Nevertheless, little information is available on the performance, under UK conditions, of coneretes containing lower binder contents than those recommended by British Standards. The experimental results reviewed in this chapter provide no empitical or scientific basis for suggesting minimum binder contents. It has thus not been felt possible to give guidance on minimum binder content in this chapter. 31Minimum requirements for durable concrete 2.9.2. Influence of curing ‘The relationship between duration of curing and rate of carbonation is not well established for curing under site conditions, Nevertheless, studies have generally shown that carbonation rates fora given type of concrete are lower for well cured specimens than for poorly cured specimens. It is generally recognized that the duration of effective curing achieved in practice is very variable. As stated in the introduction, this chapter has been based primarily upon data obtained from CEM I and CEM I/pfa concretes exposed after one day of curing and CEM V/slag concretes after three days of curing. This may be a conservative assumption, and carbonation depths for coneretes that have been subject to longer periods of effective curing will be lower than predicted. The effect of initial curing may be less significant for external sheltered exposure than for concretes subject to internal exposure, whilst for externally exposed concrete the effect may be much less. 2.9.3. Exposure classification There are significant differences between the exposure classification in pr EN 206 (BSI, 1997a) and that in BS 8110 (1997), Also, there is some difficulty in the interpretation of the carbonation, classes in pr EN 206, particularly XC3 ‘moderate humidity’ and XC4 ‘cyelie wet and dry" During development of the exposure classification it was intended that the increasing numerical designation should correspond to increasing severity. Literal interpretation of the definitions is difficult due to the absence of any quantification of ‘moderate’ humidity or the severity and frequency of the wetting and drying cycles. Concrete wets rapidly but dries much more slowly; thus in an outdoor environment in the UK the frequent rain generally means the concrete has a relatively high moisture content and consequent low rates of carbonation. Review of published information has shown that the increased time for the carbonation front to reach the reinforcement in XC4 exposure conditions more than outweighs the shorter corrosion period before damage occurs. It would thus appear that, in the UK, XC3 is a more aggressive exposure class than XC4. Although the definitions of exposure classes XC1 and XC2 are virtually opposite (ie. ‘dry’ and Swet, rarely dry’, respectively) they are both, in effect, nil classes for carbonation-induced corrosion. In the case of XC1, carbonation of the concrete may be relatively rapid but the moisture content of the concrete, by definition, will be insufficient to support significant corrosion of the reinforcement. Conversely, in the case of XC2, the rate of corrosion of the reinforcement might be quite rapid but the high moisture content of the concrete will mean that carbonation of the concrete is unlikely to reach the reinforcement within the lifetimes being considered, ie. 50 years and 100 years. Itis thus clear that the proposed pr EN 206 classification of carbonation-induced corrosion would benefit from review of the definitions of the exposure classes XCI to XC4. If the system remains unchanged in pr EN 206, it will be necessary to give guidance for use in the UK on how to interpret the classification system. ible to make a direct comparison with recommendations for conerete quality in Standards because of the different systems of exposure classification. Nevertheless, itis possible to make assumptions within the interpretation of those classifications that make broad comparisons possible. Indeed such assumptions have been necessary in the 32