Construction and Building Materials 202 (2019) 465–493

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: [Link]/locate/conbuildmat

Review

Durability of shotcrete for underground support– Review and update

Isabel Galan a,⇑, Andre Baldermann a, Wolfgang Kusterle b, Martin Dietzel a, Florian Mittermayr c
a
Institute of Applied Geosciences, Graz University of Technology, Austria
b
Faculty of Civil Engineering, Ostbayerische Technische Hochschule Regensburg, Germany
c
Institute of Technology and Testing of Building Materials, Graz University of Technology, Austria

h i g h l i g h t s g r a p h i c a l a b s t r a c t

State-of-the-art about processes and

driving forces affecting shotcrete
durability.

Focus on leaching, sintering, sulfate
attack, freeze-thaw and steel
corrosion.

Recommendations to improve
shotcrete durability.

a r t i c l e i n f o a b s t r a c t

Article history: Shotcrete is often used for the construction and stabilization of tunnels and other underground struc-
Received 19 September 2018 tures, where it is susceptible to different forms of physical and chemical attacks affecting its durability.
Received in revised form 14 December 2018 An advanced understanding of the factors that limit the durability of shotcrete is crucial to develop tai-
Accepted 21 December 2018
lored strategies for enhancing its service life. The main focus of this contribution is to shed light on the
durability of shotcrete by revising the literature, highlighting what is missing and needs to be addressed,
assessing how the knowledge about concrete durability can be transferred to shotcrete, and providing
Keywords:
recommendations for durable shotcrete structures.
Durability
Shotcrete
Ó 2018 Elsevier Ltd. All rights reserved.
Sprayed concrete
Sulfate attack
Leaching
Sintering
Corrosion
Underground
Tunnel

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
1.1. Definition of shotcrete and distinction from concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
1.2. Applications of shotcrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
1.3. Durability issues of concrete and shotcrete. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
2. Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

⇑ Corresponding author.
E-mail address: igalangarcia@[Link] (I. Galan).

[Link]
0950-0618/Ó 2018 Elsevier Ltd. All rights reserved.
466 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

2.1. The leaching process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

2.2. Controls on leaching rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
2.3. Existing models/tests to assess leaching pathways and kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
2.4. Recommendations to reduce leaching of shotcrete. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
3. Sintering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
3.1. Different forms of carbonation and sintering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
3.2. Effect of (drainage) solution composition on the nature of sintering products. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
3.3. Countermeasures for reducing the sintering potential of shotcrete. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
4. Sulfate attack and thaumasite formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
4.1. Sulfate attack in shotcrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
4.2. Role of accelerators on sulfate attack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
4.3. Other parameters controlling the sulfate resistance of shotcrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
4.4. Real damage reports in shotcrete. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
4.5. Recommendations and countermeasures against sulfate attack on shotcrete. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
5. Freeze-thaw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
5.1. Mechanisms and effects of freeze-thaw and controlling parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
5.2. Freeze-thaw in tunnels and testing methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
5.3. Recommendations to reduce freeze-thaw damage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
6. Electrochemical corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
6.1. Processes underlying electrochemical corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
6.2. Cl penetration and carbonation in shotcrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
6.3. Recommendations for reducing electrochemical corrosion of shotcrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
7. Further durability issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
7.1. Alkali-aggregate reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
7.2. Shrinkage and cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
7.3. Temperature/Fire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
7.4. Microorganisms attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
8. Final remarks and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
Conflict of interest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488

1. Introduction process requires smaller machines, and thus lower power, and
its use is needed in confined construction or inner city sites
1.1. Definition of shotcrete and distinction from concrete (e.g. metro). Because of its very rapid setting it is also used when
uncontrollable water penetration occurs.
Shotcrete, also called ‘sprayed concrete’, is a special type of con- A compilation of today’s application ranges of shotcrete, of the
crete which is applied to the substrate by shooting it at high speed ‘dos and don’ts’ of the shotcreting process and relevant recommen-
(Fig. 1). As such, shotcrete is not considered as a unique material, dations for the production of durable shotcrete are provided in e.g.
but rather as a special process to place concrete. One of the most EFNARC reports [1,2], Jolin and Beaupre [3], Kusterle et al. [4] and
important characteristics of shotcrete, specifically in tunnel appli- various national guidelines like [5–9]. With regards to the expo-
cations, is the very high early strength development, achieved in sure classes, the European Standard EN 14487-1 for sprayed con-
most cases by means of a setting accelerator, which together with crete [10] states that those specified for concrete in EN 206-1
the spraying procedure, allows concrete to be placed without extra [11] are also valid for shotcrete with 2 exceptions: (i) the recom-
support, sticking to the surface (also overheads) and hardening mended minimum cement content in the mix is 300 kg/m3, and
within minutes. (ii) the recommendation for minimum air content in concrete is
The ‘shotcreting’ process can be divided in the so called wet- not valid for shotcrete.
and dry-mix processes. In the wet-mix process (Fig. 1a-c), the
constitutive materials (aggregates, water and binder) are mixed 1.2. Applications of shotcrete
beforehand and then conveyed and pumped to the nozzle, where
the wet-mix, together with an accelerator and pressured air, is In general, shotcrete is used where shooting of the concrete mix
sprayed towards the substrate (e.g. soil, rock, tunnel walls, and is needed, that is, (i) where access is difficult and standard concrete
linings of (waste) water pipes). In the dry-mix process (Fig. 1d- casting and application is not possible and (ii) where very rapid
e) the water is added at the nozzle of the delivering hose to the setting is required. The support of rock in mining and tunnelling
flow of a humid or oven-dried pre-mixed dry mix. Both tech- is probably the most important application of shotcrete, but it is
niques have advantages and disadvantages and their suitability also used for structural repairs, refractory linings, soil stabilization
depends on the individual application. Nowadays, the wet-mix- and installation of shell structures, swimming pools, river walls,
process is used more frequently in tunnelling because of the domes and even houses. In the case of tunnels, drilled in unconsol-
much higher throughput and the lower dust production (e.g. idated sediments, soft rock or soils, shotcrete has been mainly used
reducing the subsequent workmen health risk), as compared with as a temporary ground support; in hard rock tunnelling, however,
the dry-mix process. Furthermore, the amount of rebound mate- shotcrete is often conceived as a permanent shell [12,13].
rial, comprising mainly of coarser aggregates that do not stick to The use of shotcrete for tunnelling has been developed together
the substrate and bounce off, is significantly reduced by the wet- with the tunnelling technology. Specifically, the ‘New Austrian
mix process, leading to economic benefits (e.g. reduced produc- Tunnelling Method’ (NATM), developed in the 1960s, allowed for
tion costs) and minor changes in the composition of shotcrete the first time the application of shotcrete in a totally new way,
(e.g. paste/aggregate ratio). On the other hand, the dry-mix wherein the original strength of the rock surrounding ground is
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 467

Fig. 1. Examples of shotcrete application: wet-mix process spraying in tunnels (a, b), general view of wet-mix process equipment and machinery during shotcrete tests (c),
dry-mix-process spraying (d, e).

preserved and the tunnel loads are carried by the tunnel and the (Barton et al., 1974) and the Rock Mass Rating (Bieniawski, 1976)
rock [14]. Advantages of the NATM versus the older methods fur- have been proposed. These indexes have permitted to characterize
ther include working at higher speed, lower depth of rock excava- rock masses and mountain water flow on a much higher level.
tion (smaller cross section needed), higher workplace security, etc. Noteworthy, in those times temporary layers of shotcrete sprayed
Around the same time, advances in Tunnel Boring Machine (TBM) on tunnels only needed to perform right until the permanent shell
technology resulted in a much broader use of TBMs during tun- of concrete was in place and carried all structural loads. In recent
nelling than before, and also speeding up the tunnelling progress decades, however, ongoing improvements in NATM and TBM tech-
using concrete segments; however, difficult geometries and com- nology and the target-oriented adjustment of shotcrete workability
plex (e.g. fractured and highly auriferous) body of rocks still (e.g. use of accelerating and retarding agents) and composition (e.g.
require the use of shotcrete for stabilizing and supporting exca- use of sulfate resistant cements and supplementary cementitious
vated tunnels. Since the 1980s, fibre reinforced shotcrete has been materials (SCMs)) to various natural conditions and tunnel proper-
more often applied as a permanent hard rock lining. New index ties have increased the application and service life of shotcrete to
parameters such as the NGI Tunnelling Quality (Q-method) Index 10–50 years up to 200 years in special cases [15].
468 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

A representative cross section of a tunnel structure, including - The use of different setting accelerators (e.g. alkaline
the architecture of the multi-barrier system currently used for car- aluminate-based, alkaline silicate-based, alkali-free aluminium
rying the tunnel and mountain loads as well as recent environmen- sulfate-based, etc. [18]), plasticizers (e.g. polycarboxylate
tal protection measures, is shown in Fig. 2. The degradation of ether-based superplasticizers) and cement hydration stabilis-
shotcrete may directly affect the durability of the inner concrete ers/retarders (e.g. lignosulfonates, hydroxycarboxylic and poly-
lining [16,17]. The reason for this is based on the fact that shotcrete carbonic acids and their salts) to control the pumpability of the
is in direct contact with the surrounding rock and the percolating mix and the time delay between mixing and shooting, modifies
waters and, while in good condition, it acts as a protective barrier the microstructure of the shotcrete. This, in combination with
against the penetration of potentially aggressive ionic species, such the use of fillers and SCMs (fly ash, metakaolin, microsilica,
as sulfate and chloride ions. Therefore, in double lined systems, etc.) for various purposes further influences the pathways and
impermeable technical liners are nowadays installed between the kinetics of the hydration reaction, which strongly affects the
shotcrete and the concrete layers; however, these liners may get early strength development.
brittle after several years and therefore stop providing sufficient - The characteristics of the substrate where shotcrete is applied,
protection on the long-term. In any case, once the shotcrete layer e.g. the type of rock, flow of water, and intensity of water-
is attacked and eventually destroyed, the integrity of the concrete rock interactions, control the chemical composition of the solu-
lining is put in danger. tions initially interacting with the fresh shotcrete. These,
together with the environmental conditions (temperature, rela-
1.3. Durability issues of concrete and shotcrete tive humidity, pH value, redox conditions, drying-wetting
cycles, atmospheric composition, etc.), determine the potential
Durability of concrete is a vast field of research and major con- degradation risks and deterioration rate of the shotcrete.
tributions have been made in the last decades. As concrete is the Early deformations imposed by the rock mass play also an
most commonly used material in construction all over the world, important role as they condition the adhesion between the rock
every step taken towards more durable and sustainable structures and the shotcrete and the possible formation of cracks or even
has both economic and environmental benefits. Durability of con- shear failures of the lining.
crete is always linked to exposure conditions and therefore part of - The curing of the shotcrete layers in tunnelling often relies on
the knowledge gained is widely applicable. However, some specific natural curing in the humid subsurface environment. This
issues may only be relevant to specific cases, such as exposure to may be not good enough to ensure homogeneous and complete
extremely aggressive environments. In the case of shotcrete, gen- hydration all along the thickness of the shotcrete layer, which is
eral knowledge from concrete can be used, but care has to be taken generally 20–30 cm. Insufficient curing makes shotcrete more
when adapting mixing designs, properties, testing methods and susceptible to chemical and physical attacks, with the outer
guidelines values. In particular, the following aspects, specific for shotcrete layers typically presenting a higher porosity and
shotcrete, have to be considered: lower amounts of hydration products than the bulk.

- The shotcreting process determines the compaction degree and Shotcrete used in underground structures, like tunnels, is in
thus the microstructure (e.g. pore size distribution and perme- direct contact to the surrounding rocks and the corresponding per-
ability) of the final shotcrete. Some parts of the shooting pro- colating waters. Depending on the characteristics of the natural
cess are automated and robot operated, while many other interacting solutions (e.g. pH, ionic strength and element concen-
parts depend on the experience of the nozzle man, who controls tration, temperature, and flow rate), susceptibility to dissolution
the shooting equipment. The settings and adjustments of the and/or chemical leaching can be very high, especially in contact
shooting parameters, such as the compressed air pressure or with low mineralized solutions. The early dissolution of solid
the accelerator dosage, have a direct impact on the homogene- phases, such as portlandite (Ca(OH)2) and calcium-silicate-
ity and final quality of the sprayed layers. hydrate (C-S-H), originally present in the shotcrete can cause

Fig. 2. Double-shell tunnel cross section. The architecture of the multi-barrier system, the environmental protection measures and the critical zones of physical and chemical
‘attacks’ affecting the durability are indicated. On the left side of the figure the potential durability issues are highlighted together with the chapter in which the topic is
discussed.
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 469

dramatic changes in the composition and ionic concentration of In the last decades, little emphasis has been placed on the iden-
the interstitial solutions. In the case of sulfate attack interstitial tification and quantification of the reaction paths, mechanisms and
solutions frequently show extremely high oversaturation degrees environmental controls underlying shotcrete deterioration. Until
with respect to hydrous calcium sulfates (e.g. gypsum (CaSO42H2- recent, the durability of shotcrete was thought to be mainly depen-
O), ettringite (Ca6[Al(OH)6]2(SO4)326H2O) and thaumasite (Ca3Si dent on its permeability [19] and emphasis has been mainly put on
(OH)6(CO3)(SO4)12H2O)). In contrast, calcium carbonates (CaCO3) the technical and practical aspects [20]. The complexity and the
(e.g. calcite, aragonite, vaterite) tend to precipitate in the drainage impact of multi-faceted physical and chemical attacks have thus
system of tunnelling structures in the course of leaching and sin- been often ignored when predicting the long-term performance
tering of shotcrete (Fig. 3). of shotcrete under different environmental conditions. The highly

Fig. 3. Examples of durability issues in shotcrete structures: sulfate attack, thaumasite formation and efflorescence in tunnels (a-d), carbonate sintering in tunnel drainage
systems (e, f).
470 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

complex interaction mechanisms between shotcrete and (intersti- through a porous cementitious material, such as shotcrete. This
tial) solutions typically result in a suite of chemical attacks (Fig. 2), process can result in alteration of system equilibria and decompo-
with leaching (chapter 2), sintering (chapter 3) and sulfate attack sition of phases, which subsequently lead to an increase of poros-
(chapter 4) being the most important ones in terms of the technical ity, reduction of strength, and finally disintegration of the shotcrete
and economic difficulties they cause. Besides these attacks, there (Fig. 4a, b, c). A schematic representation of a leaching process is
are other processes that can affect the lifetime and performance shown in Fig. 4d.
of shotcrete. Additional important issues covered in this review The interaction between so-called soft waters, i.e. poorly miner-
include exposure to freeze-thaw cycles (chapter 5), electrochemi- alized natural solutions (e.g. meteoric and drainage solutions), and
cal steel corrosion (chapter 6) and other durability related issues ordinary Portland cement (OPC) dissolves portlandite (Eq. (1)) and
like the alkali-aggregate reaction, the resistance to fire and high leaches alkalis and hydroxide ions, subsequently creating strong
temperatures, the dimensional stability and shrinkage, and the pH gradients at the water-shotcrete interface. After portlandite
resistance to microorganisms (chapter 7). The most important dissolution has been almost completed and the pH of the pore
issues affecting the durability of shotcrete are described and dis- solutions is getting lower, C-S-H phases are progressively decalci-
cussed. Currently used standards and guidelines, methods and data fied and decomposed (Eq. (2)), as are the aluminate phases
published in the literature, together with recommendations, are [21,22]. All this leads to the development of microcracks in the
provided. damaged zone and in an initial chemical weakening of the cement
paste, so that in the final stages solely hydrated silica gel, alumina
2. Leaching and iron oxides remain [23]. As a result of ongoing leaching and
diffusive/advective transport of the dissolved species at the active
2.1. The leaching process leaching front, high concentration gradients are created in the pore
solutions, in turn promoting mineral precipitations in the ‘outer
Leaching is a process where a solid mixture dissolves, usually parts’ of the leached shotcrete and further dissolution action in
the ‘inner parts’. The sequence and extent of dissolution and pre-
resulting in the formation of an inert or less soluble solid residue
and a solute component, i.e. ions and uncharged aquo- cipitation reactions depend mainly on the composition of the
cement paste and of the solvent. For instance, in a standard OPC
complexes. Thus, a typical leaching operation occurs when an
aqueous solution undersaturated with respect to (highly reactive) mix, leached by soft water, portlandite dissolves first, followed
by AFm (Ca4Al2(SO4)(OH)126H2O), ettringite and calcite. Secondary
cement phases, such as portlandite and C-S-H, passes over and/or

(a) (b)

(c) Sample surface (d)

Air
void

Cement
paste

Aggregate

200 µm

Fig. 4. Examples of leaching in shotcrete: Aggregates leached from a sample exposed to soft water and low temperature (a) and specimens exposed to weak acidic attack
(b). Microstructure of leached surface (c). Scheme of leaching front progression towards the sound zone (d).
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 471

AFm forms where portlandite dissolves, and secondary ettringite mainly on the pH and presence of metal ion aquo-complexes. For
precipitates during the dissolution of portlandite and AFm [24]. example, the dissolved carbonate ion (CO2– 2+
3 ) can react with Ca ,
Similar alteration features have recently been described by Rosen- e.g. from Ca(OH)2 and C-S-H dissolution, to form the uncharged
qvist et al. [25], who report on the alteration of concrete after (CaCO3)0 complex, from which the CaCO3 polymorphs, i.e. calcite,
55 years of exposure to (soft) river water: decalcification, forma- aragonite and vaterite, can precipitate (Eq. (8)). However, through-
tion of secondary ettringite and precipitation of a magnesium- out the leaching process, CaCO3 phases can be dissolved, creating
rich silica gel constituted the major chemical and mineralogical connected pores and fluid migration paths. In this case, the term
changes. ‘aggressive CO02 is used to delimitate the amount of CO2 dissolved
in the water, where water rich in CO2(aq) speeds up the apparent
CaðOHÞ2 ðsÞ ! Ca2þ þ 2OH ð1Þ leaching rates. The reaction between concrete and natural waters
containing CO2(aq) was described by Cowie and Glasser [40], who
C-S-H ðsÞ ! xCa2þ þ 2xOH þ SiO2  nH2 O ð2Þ developed a model based on solubility and thermodynamic data
to create aggressivity maps with leachability contours (i.e. ratio
Aggregates may be affected by leaching in two ways: (i) physi-
of cement substance dissolved relative to that dissolved by pure
cal outbreak through a loss of bond strength between paste and
water) and also maps including precipitation areas of CaCO3.
aggregate as a consequence of the surrounding paste being leached
and decomposed (Fig. 4b), and (ii) leaching and dissolution of the CO2 ðgÞ ! CO2 ðaqÞ ð5Þ
aggregates themselves (Fig. 4a). For instance, the risk for dissolu-
tion of calcite- and dolomite (CaMg(CO3)2)-based aggregates CO2 ðaqÞ þ OH ! HCO3 ð6Þ
(Eqs. (3) and (4)) increases when exposed to soft or acidic water
at low temperature where its solubility increases. Examples of this
HCO3 ! CO2
3 þH
þ
ð7Þ
type of aggregate leaching are found in head race tunnels of hydro
power plants. However, also under strongly alkaline conditions (i.e.
pH range from 13 to 10.5), provided through the dissolution of Ca2þ þ CO2
3 ! CaCO3 ðsÞ ð8Þ
portlandite and C-S-H, dolomite was found to dissolve rapidly, pro- Most authors coincide on the positive effect of reactive addi-
moting the subsequent formation of calcite, brucite (Mg(OH)2) and tions, such as silica fume and fly ash, because they consume cal-
thaumasite [17]. From these examples one can conclude that cium, yielding in less portlandite, during cement hydration. Less
leaching of shotcrete often marks the beginning of the other types portlandite in the shotcrete/concrete mix reduces its leachability
of chemical attacks, such as sulfate attack [26–29], alkali-aggregate [41]. Accordingly, denser matrices are formed by the addition of
reactions [30–33], and sintering/carbonation [34–36]. silica fume, fly ash and/or granulated blast furnace slag, reducing
CaCO3 ðsÞ ! Ca2þ þ CO2 ð3Þ the penetration of concrete by aqueous solutions [42]. On the other
3
hand, portlandite, maintaining the pH in the pore solution, delays
the access of acids to the cement matrix by ongoing neutralisation
CaMgðCO3 Þ2 ðsÞ ! Ca2þ þ Mg2þ þ 2CO2
3 ð4Þ
reactions [43]. Specifically for shotcrete, the composition of the
Leaching may also take place during shotcrete hardening. How- accelerators and retarders used may also play an important role
ever, due to the very fast hardening, the consequences of leaching in the leaching process. For example, alkali-containing accelera-
at this stage are considered insignificant both for the shotcrete tors, nowadays barely in use due to their health and environmental
itself and for the drainage solution generated [37]. In the case of risks, (i) contribute to the total leachable alkalis that could then
tunnels, the drainage system is installed some days/weeks after take part in an alkali-aggregate reaction and (ii) trigger carbona-
shotcreting and it can be easily cleaned during the ongoing tunnel tion by creating highly alkaline solutions during dissolution; pro-
construction. cesses that could decrease the durability of shotcrete.

2.2. Controls on leaching rates 2.3. Existing models/tests to assess leaching pathways and kinetics

In general, rates of dissolution and leaching are strongly con- Several models have been proposed to describe the leaching of
trolled by the component-specific solubility, the permeability of concrete, many of them considering an unaltered shrinking core
the shotcrete under attack, the water flow rate, the chemical com- and a moving leached shell. Mainguy and Coussy [44] used a
position of the solvent (e.g. pH and ionic strength) and the temper- dissolution-diffusion approach considering a sharp leaching front
ature. The presence of acids usually increases the rate of chemical that separates a completely dissolved zone from a non-degraded
attack by reacting with and dissolving the alkaline constituents of one (Fig. 4c,d). The model takes into account that, for leaching, dif-
hydrated cement and certain aggregates, such as limestone [38]. fusion of ionic species through a porous medium (e.g. concrete) is
Acids that can attack cementitious materials are generated in var- nonlinear because of the dependence of the effective diffusion
ious natural and man-made systems, including groundwater and coefficient on the changing porosity, and that dissolution happens
acidic soils (sulfuric, carbonic, carboxylic, humic, lactic), agricul- instantaneously. The Nernst-Planck equations have also been used
tural and urban areas (lactic, butyric, acetic), acid rain (carbonic, to model leaching and penetration of ions in concrete [45], with a
sulfuric, nitric), effluents from industrial processes and sewers further verification of data by electron microprobe analysis. Glas-
(sulfuric), and seawater (carbonic) [38]. Attack by strong acids ser et al. [46] reviewed the mechanisms of ionic transport in con-
has recently been reviewed by Duchesne and Bertron [39], con- crete and the models proposed in the literature for the
cluding that the severity of the attack is predominantly dependent decalcification process: the simplified ionic transport equation
on the solubility of the calcium salt of the respective acid, like Ca- mostly used neglects the influence of common phenomena such
tartrates, Ca-carbonates and Ca-sulfates, and the stability of the as chemical activity, convection and electrical coupling.
corroded layer under the given conditions. Some recent contributions include thermodynamic equilibrium
The most frequent and important acid found in natural water is calculations coupled with 3D cement paste microstructure models
carbonic acid, forming from the dissolution of gaseous CO2 in water [47]. In this approach, the leaching process is divided in three
(Eqs. (5–7)). In aqueous solutions, four dissolved inorganic carbon stages: (i) dissolution of portlandite until its full consumption,
species can be found: CO2(aq), H2CO3, HCO–3 and CO2– 3 , depending (ii) reduction of Ca/Si ratio of C-S-H, and (iii) dissolution of
472 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

ettringite and hydrotalcite, leaving an amorphous silica remnant as are two specific methods to assess the leachability of shotcrete
the only solid phase at the end of the reaction. ‘If the exchange [9]. Both approaches are based on the immersion of shotcrete spec-
between the leaching solution and the interior is slow enough for imens in water. For the first method 28 days old shotcrete drilled
the pore solution to reach equilibrium with the solid phases, the cores are placed in a container with deionised water (shotcrete:
dissolution of ettringite and the remaining calcium in C-S-H can water mass ratio 1:10) and continuous stirring. After 24 h the pH
promote a transient driving force for precipitation of gypsum’ of the solution, electrical conductivity, and concentration of dis-
[47]. Other approaches to describe and predict the leaching process solved calcium, sodium, potassium and aluminium are measured.
in concrete with the use of thermodynamics, assuming equilibrium The second method, the so-called RV test [86] is used to evaluate
conditions between the pore solution and the hydrated phases, the sintering potential of shotcrete and it is described in Section 3.
include those from Jacques et al., Stora et al., Engelsen et al., Gerard
et al., Hidalgo et al. and Nakari et al. [48–53]. Hidalgo et al. [52] 2.4. Recommendations to reduce leaching of shotcrete
state that the coupled dissolution of portlandite, decalcification
of C-S-H, silicate polymerisation, and incorporation of tetrahedral In most double-lined tunnels, leaching of concrete layers is pre-
(and/or octahedral) coordinated aluminium in the polymerised vented by the so-called ‘umbrella solution’, which is based on the
gel, result in the formation of an aluminosilicate gel (C-A-S-H) with use of membrane-insulation-systems situated between the shot-
higher surface area, higher internal porosity and higher interlayer crete and concrete inner lining (Fig. 2). These insulation systems,
water content than the initial C-S-H gel. De Larrard et al. [54] which allow for the drainage of the water, generally consist of a
developed a model for calcium leaching from concrete exposed geotextile layer and a waterproof polymer sheet membrane made
to water based on the mass balance equation for calcium and of PVC or PE [87]. In recent years, sprayed membranes are being
implemented in a finite volume numerical code, which was vali- tested, especially for permanent sprayed concrete linings [13],
dated with literature data. In a parametric survey the authors but due to the lack of regulations and the uncertainty in the long
found that the most influential parameters on the leaching kinetics term bond performance its use is still very restricted [88]. In some
are the initial ‘water porosity’ of the materials (i.e. porosity mea- cases, where no membrane is used, watertight inner linings are
sured by immersion in water under vacuum plus drying at installed. The insulation layer can also comprise a full-round
105 °C) and the effective aqueous diffusivity of calcium ions watertight membrane but this is a less common method. Some
(Ca2+) within the pore structure. Gawin et al. [55–57] proposed tunnels and special constructions, especially in Europe, also use
the use of an equation describing kinetics of the process instead the so-called ‘Scandinavian method’ in which the percolation of
of an equilibrium curve, taking also into account the advective flux groundwater is prevented by a permanent shotcrete protection
of Ca2+ through a porous cementitious structure. shell [89]. However, the shotcrete layer, in direct contact with
The above discussed leaching pathways are linked with pro- the rock, remains exposed to leaching for its entire service life irre-
gress in mechanical deterioration of the concrete, often described spective of all recently available technical countermeasures.
by chemo-mechanical models [58–68]. Le Bellego et al. [67] The addition of SCMs to the cement to lower the amount of
reported a decrease of stiffness, strength and fracture energy and portlandite that can form upon hydration is a possible path to fol-
an increase in brittleness of leached concrete. Kuhl et al. [61] low to reduce leaching in shotcrete structures. The use of SCMs
showed that chemical damage reduces the limit load, the stiffness such as silica fume, fly ash and metakaolin has also environmental
of concrete and the structural behaviour. Carde and François [68] and economic benefits, but the new shotcrete mixes have to be
reported a decrease in the local elastic modulus with a damage optimized, so that the early strength gain development is not com-
function and an aging function, this latter dependent on the cal- promised [90]. Such additions can further improve the leaching
cium content, which in turn changes with the thickness of the resistance, forming denser matrices and hindering the penetration
degraded zone and the different types of cement hydrates. On of shotcrete by aqueous solutions. Yoshida et al. [265,266] also
the other side, cracking also plays a role on the transport processes reported on the beneficial use of granulated blast furnace slag to
in concrete [69,70], including leaching: Gerard and Marchand [71] improve the resistance of shotcrete to sulfuric acid leaching. The
reported on the marked influence of continuous cracks on the con- authors measured the specimens’ mass change and neutralization
crete transport properties, especially for denser matrices where the depth after immersion in sulfuric acid solutions (5 and 0.1%) and
effect is relatively more important. observed major differences in the leaching behavior when compar-
Experimental leaching tests for concrete include static and run- ing OPC- and granulated blast furnace slag cement-based speci-
ning water [72,73], exposure to weak (pH 6.5–5.5), strong (pH 5.5.- mens exposed to the higher concentration of sulfuric acid at
4.5) and very strong (pH < 4.5) acidic solutions, adjusted by contin- 20 °C. The accelerator used also played an important role: samples
uous addition of diluted acetic, hydrochloric, sulfuric or nitric acid with 7% of a calcium aluminate type accelerator behaved better
[74,75], and accelerated methods (i) where electrical fields are than those with 10% of a calcium sulfoaluminate type. The amount
applied to facilitate diffusion ion transport within the pore solution and distribution of portlandite in the mixes, the volume of the
[76–78] or (ii) aggressive ammonium nitrate solutions are used to reaction products and the variation in porosity are mentioned as
speed up the dissolution rate of cement-based materials [79–81]. possible explanations for the different behavior observed; how-
The range of specimen sizes used in the tests varies from a few ever, no further evidence is given.
millimetres to several tens of centimetres. Some authors used The composition of the accelerators, stabilizers, plasticizers and
powdered samples to analyse maximum leaching potential by retarders used in shotcrete and their interaction with the cement
removing the influence of pore structure or diffusion paths paste and aggregates play a key role in controlling the ongoing
[50,52,82,83]. Some of these tests [84] are based on the Dutch leaching process (direction, mechanism, pathways, kinetics, etc.);
standard NEN 7341, where the sample has to be ground yet the interplay between these components is still poorly under-
to <125 mm to increase the surface area and reactivity, and subse- stood. A proper selection of aggregates used in shotcrete can help
quently a two-step extraction (i.e. exposure to solutions at pH 7.0 to increase its resistance against physical and chemical corrosion.
and 4.0 for 3 h) is applied at a liquid to solid ratio of 50 l/kg [85]. The applicability of the various aggregates such as limestone, dolo-
The results of the leaching tests are influenced by the applied mite, sand and gravel, depends mainly on the physicochemical
physicochemical boundary conditions. The use of a standard test conditions prevailing in the tunnel which in turn are controlled
method that best replicates the leaching conditions in the field is by the intensity of water-rock interactions and related weathering
still hotly debated. Apart from the tests used for concrete, there paths.
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 473

According to EFNARC [2] if shotcrete is exposed to high pres- CaCO3 precipitates have to be removed from the drainage system
sure, the water permeability should be measured to assess the to ensure its function, preventing blockages and water pressure
shotcrete’s resistance to leaching. In low pressure settings, such increase; chemical and/or mechanical remove of the sinter from
as natural atmospheric exposure, the capillary suction of the cover the drainage and receiving streams requires a well-defined long-
zone of the shotcrete [267] should be measured. When water-tight term strategy for repairing and maintenance works and thus high
shotcrete is required for a highly aggressive exposure class, the expenses.
maximum value of penetration, in accordance with ISO 7031 According to Kusterle et al. [86] the chemical composition of the
[268], should be 50 mm and the water permeability <1012 m/s. water penetrating the system, the tunnel layout and the cement or
The Austrian Guideline for Sprayed Concrete [9] gives recommen- binder composition are the main influential factors controlling the
dations for shotcrete exposed to one of the three classes specified sintering process, although other factors, like water flow rate, shot-
in the European standard EN 206-1 [11]. The exposure classes XA1, crete permeability and moisture content, presence or absence of
XA2 and XA3 for low, moderate and high risk of leaching, respec- chemical inhibitors and temperature are also important [91]. The
tively, are classified in terms of the pH value and the concentration pathway, mode and rate of CaCO3 precipitation in tunnels, con-
of CO2(aq), NH+4 and Mg2+ ions in the groundwater. For shotcrete trolled by the interplay between the shotcrete, the local percolat-
exposed to XA1 a dense matrix is required: this should be ensured ing water and the atmosphere, may be traced by analysing the
by a penetration depth, as measured by the ÖNORM B 3303 [269], stable oxygen and carbon isotopic composition of the precipitating
lower than 50 mm. When exposed to XA2 the maximum penetra- carbonates [93,100].
tion depth is reduced to 35 mm. In this case the finer aggregates, Postulated mechanisms of carbonation and sintering are visu-
<4 mm, should not contain carbonates in percentages higher than alised in Fig. 5 and include (i) degassing of CO2 from the solution
15% CO2 equivalent, except when renewal of the liquid is negligi- in the atmosphere, (ii) mixing of different water types, (iii) increase
ble, where calcitic and dolomitic aggregates should be used to neu- of pH due to dissolution of portlandite (or other alkaline hydrox-
tralise the solution. Finally, the exposure class XA3 requires special ides) and (iv) absorption of CO2 at high pH, according to Eqs. (9)–
measures and sprayed concrete should not be used as a permanent (12) [92].
support under those conditions. The dominant precipitation regime apparently controls the
characteristics of the carbonate deposits: either dense or weak
3. Sintering and porous microfabrics build up by calcite crystals (Fig. 5e) can
form through sintering of drainages with or without direct contact
3.1. Different forms of carbonation and sintering to shotcrete. More rarely, deposition of yellowish to reddish-brown
Fe-rich layers, comprising mainly of highly reactive Fe-O-OH and/
Carbonation of shotcrete refers to the chemical process or Mn-O-OH precipitates, e.g. ferrihydrite (Fe2O30.5H2O) and
whereby portlandite and/or C-S-H react with atmospheric CO2 in goethite (FeOOH), accompanies sintering. This is in most cases
an aquatic surrounding to form CaCO3 and silica gel (SiO2nH2O) related to microbial processes (see chapter 6) and/or intense
as a reaction by-product (Eqs. (9–12)). During this neutralization weathering of Fe(II)-bearing minerals (i.e. iron sulfides) from the
reaction, the pH of a fresh cement paste changes from 12.6 or rock which results in the mobilization of Fe(II) species and in the
higher to around 8 if completely carbonated. Carbonation fre- subsequent formation of Fe(III)-(oxy)hydroxides.
quently results in a densification of the microstructure due to a
decrease of the porosity, because of the larger volume occupied 3.2. Effect of (drainage) solution composition on the nature of sintering
by CaCO3 compared to portlandite, which can be an advantage in products
non-reinforced shotcrete. In reinforced shotcrete, however, it can
be a great drawback: if the carbonation front reaches the steel, Apart from quantitatively less important mineral precipitations
its passivation regime shifts to corrosion due to pH decrease. The on the tunnel walls, massive and fast blocking/clogging of drai-
progress of the carbonation front from the shotcrete surface nages by sintering reactions is a major issue in tunnelling. Drainage
towards the steel reinforcement controls the corrosion risk. This solutions in tunnels – the probably most important trigger for sin-
aspect will be further described in Section 6 together with chloride tering – originate from the interaction between local percolating
penetration in relation to electrochemical corrosion. solutions, most likely groundwater, and shotcrete/concrete layers.
Typical chemical reaction pathways of carbonation are given As it is indicated early in this chapter, this interaction leads in
for calcium carbonate formation by the following equations many cases to the massive precipitation of carbonate minerals,
(Eqs. (9–12)): with calcite being the most abundant CaCO3 phase, precipitating
directly at the shotcrete/concrete layer and/or subsequently within
Ca2þ þ 2HCO3 ! CaCO3 þ H2 O þ CO2 ðgÞ ð9Þ the drainage system. The chemical composition of the drainage
solutions is a key controlling the kind of precipitating sinter
CaðOHÞ2 þ Ca2þ þ 2HCO3 ! 2CaCO3 þ 2H2 O ð10Þ (Fig. 5). In the ‘normal’ mode of CaCO3 precipitation (Eqs. (9–12))
calcite is the main sintering product, but the anhydrous CaCO3
polymorphs aragonite or vaterite [93,94] or, under special low
Ca2þ þ 2OH þ CO2 ðgÞ ! CaCO3 þ H2 O ð11Þ
temperature conditions, amorphous CaCO3 (CaCO3nH2O) and
ikaite (CaCO36H2O) can also be formed [95,96]. Most decisive
C  S  H þ xCO2 ðgÞ ! xCaCO3 þ SiO2  Hn2 O ð12Þ
parameters for the formation of sinter are the concentration levels
Another phenomenon related to carbonation reactions, espe- of individual dissolved components (e.g. Ca2+, dissolved inorganic
cially relevant in tunnel structures, is the so-called sintering or carbon, Mg2+, presence/absence of inhibitors), pH, CO2 partial pres-
scaling (note here that these terms are not exclusively used for car- sure (pCO2) and temperature. For example, aragonite typically
bonate deposits). In this case, with the pre-requisite of calcium forms at molar Mg/Ca ratios >2 at ambient conditions instead of
leaching of shotcrete and concrete and/or calcium supply from calcite (and vaterite); the presence of polyaspartic acid (a polymer-
groundwater, precipitation of massive CaCO3 minerals takes place ized amino acid which is used to prevent carbonation in drainages)
on the tunnel walls and/or in the tunnel drainage system (Fig. 5a, b, can inhibit aragonite formation and favour vaterite precipitation; a
d). This latter has severe technical and economic consequences: combination of polyaspartic acid and magnesium salt addition can
474 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

(a) (b)

(c) (d)

(e) (f)

Fig. 5. Carbonate sintering on tunnel walls (a, b) and drainage system (d). Sinter formation in drainage pipe (c, e). Scheme of sinter growth (f).

induce calcite formation even at high supersaturation degrees of Rinder et al. [100] aimed at elucidating the processes and reac-
the solution with respect to this mineral phase [97]. The latter tion pathways underlying sinter formation at 3 tunnel sites in Aus-
may also promote the formation of amorphous calcium carbonate tria using an integrated elemental, isotopic and modelling
and ikaite instead of the anhydrous CaCO3 polymorphs, but from a approach. Their work indicated that the physicochemical interac-
thermodynamic point of view these temporally occurring sinters tion mechanisms between different water types and shotcrete/-
are unstable and thus they will transform to the anhydrous coun- concrete, and the intensity of water-rock interaction in general,
terparts on the short-term (e.g. within minutes to days), and finally can be traced by a positive correlation between dissolved K+ and
to the thermodynamic stable polymorph calcite under ambient pH (from concrete leaching) and a decrease of aqueous Ca2+, dis-
conditions [95,98]. solved inorganic carbon and Mg2+ due to the formation of calcite
Three types of waters are commonly considered as drainage and brucite. The internal pCO2value is proposed as a suitable proxy
waters in tunnels: (i) pristine groundwater, with a chemical com- to evaluate the development of calcite saturation in the drainage
position predefined by the interactions with the local rocks, (ii) solution caused by CO2 absorption in alkaline solutions. Above a
drainage solutions with pH from 9 to 11, and (iii) drainage waters pCO2 threshold value uptake of CO2 from the air will lead to an
with very high pH, from 11 to 13 [99], resulting from moderate to increase of the saturation degree of the drainage solution with
very intense interaction with the shotcrete. The ‘high pH’ water is respect to calcite (which is commonly the case for drainage solu-
commonly enriched in Ca2+, Na+, K+ and Cl ions, especially with tions with a pH above 11), and vice versa. In essence, increased sin-
respect to the groundwater, but typically depleted in aqueous tering in tunnels and in their drainage systems is mostly caused by
Mg2+, SO24 and dissolved inorganic carbon [93]. The relative deple- the contact of local groundwater with shotcrete/concrete. For a
tion of the latter dissolved components can be used as an indicator given tunnel the individual local boundary conditions in respect
for the formation of other sinter products such as brucite, gypsum, to water quantity and quality have to be assessed in order to
ettringite and various types of carbonate minerals mentioned develop proper strategies for tailored tunnel construction features
above. and shotcrete composition.
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 475

3.3. Countermeasures for reducing the sintering potential of shotcrete tions between sulfate ions and hydrated calcium aluminates orig-
inally present in the fresh cement paste (Fig. 6). This study only
Apart from the use of blended cements to reduce the leachabil- covers External Sulfate Attack, while Internal Sulfate Attack and
ity and thus the sintering potential of shotcrete/concrete liners Delayed Ettringite Formation are not within the scope. External
(Section 2), the reduction of mixing of groundwater with highly Sulfate Attack in shotcrete frequently occurs as a result of the
alkaline solutions is recommended to prevent carbonate forma- interaction of SO2
4 ions from local groundwater or drainage solu-
tion. In this respect, Rinder et al. [100] suggested to increase the tions with the hydrated cement components. Besides the chemical
CO2 exchange rate between the atmosphere and drainage solutions composition and concentration of the sulfate-containing water, the
having internal pCO2 values between 103.45 and 106.70 atm at permeability of the shotcrete and the composition of the cement,
12.5 °C, e.g. ventilating with air, and to decrease the respective in particular the aluminate content, are key factors involved in sul-
pCO2 exchange rate for solutions with internal pCO2 val- fate attack. Also, the environmental conditions, such as relative
ues <106.70 and >103.45 atm at 12.5 °C, e.g. using laminar flow. humidity, pCO2 and temperature, determine whether sulfate attack
These measures help to decrease the saturation degree of the drai- would take place and to what extent.
nage solutions with respect to CaCO3 phases, thus inhibiting mas- The most abundant sulfates involved in the process are those of
sive and fast precipitation of CaCO3 in the affected tunnel sections. calcium, sodium and magnesium. Sodium sulfate (Na2SO4) reacts
To follow the individual reaction paths and rate of sinter formation with portlandite to form gypsum, thereby releasing Na+ and OH–
in a certain site Dietzel et al. [94] recommend a tunnel-specific ions (Eq. (13)). Magnesium sulfate (MgSO4) decomposes the C-S-H
hydrochemical monitoring campaign. Suggestions for preventing gel and reacts with the aluminates and portlandite, forming gyp-
scaling further include: optimized construction of the drainage sum, brucite, and magnesium silicate gel (M-S-H) (Eq. (14)). As bru-
system (separate or combined discharge system), use of inhibitors, cite forms, pH decreases in the cement matrix due to an anion
and structures monitoring [91]. exchange: OH– is removed from the pore solution and precipitated
Two types of inhibitors are currently being used to prevent car- in brucite, SO2
4 enters the matrix. Calcium sulfate (CaSO4) reacts
bonate precipitations in primary and secondary tunnel drainage with the hydrated aluminates to form ettringite (Eq. (15)). Accord-
systems: liquid (polyaspartic acid) and solid (polysuccinimid- ingly, in most cases, fast progressing sulfate attack results in the for-
based) [92]. According to Niedermayr et al. [97] the inhibition mation of ettringite and gypsum, combined with the dissolution of
effect of polyaspartic acid on CaCO3 formation is based on its portlandite and leaching of C-S-H. Because of the volume increase
adsorption on reactive surfaces or nucleation and growth sites of that results from the formation of gypsum and ettringite, in some
CaCO3. In the experiments reported even low concentrations of cases up to double, this type of attack typically leads to expansion,
polyaspartic acid (0.1 mg/l) led to significant retardation of CaCO3 cracking, strength loss and finally complete disintegration of the
nucleation, further enhanced if aqueous Mg2+ was present. shotcrete (Fig. 6c, e, g). The theories proposed to explain the expan-
Due to the complexity of the reactions involved, advices to sion mechanism upon precipitation of ettringite have been recently
reduce the sintering potential and/or general assertions about the reviewed [102], the most accepted one being the ‘crystallization
medium- to long-term (e.g. years to tens of years) behaviour of pressure theory’, which relates the pressure exerted by a crystal pre-
tunnel systems should be evaluated individually for each site. cipitated from a supersaturated solution to the supersaturation
Recently, the development and installation of so-called sinter ratio (i.e. ion activity product/equilibrium solubility product)
guards in affected tunnels has become a highly promising tool to [103]. Müllauer et al. [104] found out that ettringite is especially
monitor the carbonation/sintering potential on-site. Due to the harmful when formed in small (capillary) pores (10 – 50 nm).
specific electric conception, i.e. a combination of electric conduc-
tivity measuring cells, sinter guards are able to monitor the sinter
CaðOHÞ2 þ Na2 SO4 þ 2H2 O ! CaSO4  2H2 O þ 2NaOH ð13Þ
formation potential alongside a distinct water flow path in drai-
nage systems; yet their long-term stability and their robustness CaðOHÞ2 þ MgSO4 þ 2H2 O ! CaSO4  2H2 O þ MgðOHÞ2 ð14Þ
against internal CaCO3 precipitation has to be proven.  
To quantify the leaching/sintering potential of a sprayed con- Ca4 AlðOHÞ6 2 ðSO4 Þ  6H2 O þ 2ðCaSO4  2H2 OÞ þ 16H2 O
crete Kusterle et al. [86] proposed a test method called ‘RV test’:  
! Ca6 AlðOHÞ6 2 ðSO4 Þ3  26H2 O ð15Þ
after shooting the shotcrete in 50 x 50 x 15 cm boxes and leaving
to cure, 50x100 mm cylindrical specimens are wet-drilled after Thaumasite sulfate attack is a special type of sulfate attack by
56 days. With a water/specimen ratio of 4 the specimens are which the cementitious material is transformed into a mushy pro-
placed in sealed containers and left for 24, 48 and 120 h (3 cycles). duct, thereby completely losing its binding/mechanical properties.
After each cycle the fluid phase is renewed, and pH and electric The consequences of this type of sulfate attack can be more severe
conductivity are measured in the reacted solutions immediately than those of the other types mentioned before (Fig. 6a, b, d). Low
after sampling. The evolution of the Ca2+ concentration, deter- temperatures, high relative humidity and sources of silica and car-
mined by acidification to yield a pH of 3–4, defines the leaching/ bonate are required for the reactions to proceed (Fig. 6h). As silica
sintering potential of the test specimen, expressed as kg of Ca2+ sources are typically provided from the alteration of C-S-H, sources
per ton of sprayed concrete. From recent and past experience, of dissolved inorganic carbon can originate from groundwater,
wet-mix shotcrete with a maximum CEM I content of 280 kg/m3 limestone/dolomite fillers [105–114], limestone aggregates [115],
mixed with 140 kg/m3 of a combination of fly ash, granulated blast and eventually atmospheric CO2 [116]. According to Köhler et al.
furnace slag and limestone, will give RV values of 0.7 kg/to or lower [117] thaumasite preferentially nucleates heterogeneously on the
[87,101]. surface of ettringite, due to the structural similarities of the two
minerals, followed by epitaxial growth of thaumasite. Thaumasite
formation is usually slowly progressing and commonly initiated by
4. Sulfate attack and thaumasite formation cracking and softening due to secondary gypsum and ettringite for-
mation, followed by a solid solution between ettringite and thaum-
4.1. Sulfate attack in shotcrete asite, the so-called woodfordite route (Eq. (16)) [17]. Other
possibilities for thaumasite formation are the direct precipitation
Sulfate attack in shotcrete comprises a highly complex from oversaturated pore solutions or the reaction of CaCO3 with
sequence of coupled leaching, dissolution and precipitation reac- severely Ca-leached C-S-H (Eqs. (17a), 17b) [103,118–121].
476 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

Fig. 6. Thaumasite (Tha) sulfate attack and de-dolomitization (brucite (Brc) formation) in Austrian tunnels (a, b, d, f). Damaged mortar prisms after 200 days immersed in
30 g/L Na2SO4 solution at 20 °C (c). Scheme of ettringite (Ett) and gypsum (Gp) formation under accelerated test conditions (g). Scheme of replacement of cement paste and
dolomite (Dol) aggregates by thaumasite (h).

 
Ca6 AlðOHÞ6 2 ðSO4 Þ3  26H2 O þ Ca3 Si2 O7  3H2 O þ CaCO3 Ca3 Si2 O7  3H2 O þ 2ðCaSO4  2H2 OÞ þ 2CaCO3 þ 24H2 O
 
þ CO2 þ xH2 O ! Ca6 SiðOHÞ6 2 ðCO3 Þ2 ðSO4 Þ2  24H2 O þ CaðOHÞ2 ð17aÞ
 
! Ca6 SiðOHÞ6 2 ðCO3 Þ2 ðSO4 Þ2  24H2 O þ CaSO4  2H2 O
Ca3 Si2 O7  3H2 O þ 2ðCaSO4  2H2 OÞ þ CaCO3 þ CO2
þ Al2 O3  xH2 O þ 3CaðOHÞ2 ð16Þ  
þ 23H2 O ! Ca6 SiðOHÞ6 2 ðCO3 Þ2 ðSO4 Þ2  24H2 O ð17bÞ
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 477

The ratio SO3/Al2O3 plays a key role in the formation of thaum- from 2 to 10% [131] for testing the sulfate resistance of shotcrete
asite versus ettringite. According to Schmidt et al. [122] molar under accelerated conditions (Fig. 6c). In general, more inhomoge-
ratios SO3/Al2O3 above 3 are required for thaumasite to form, and neous matrices have been reported for shotcrete as opposed to
the aluminium present must have been previously consumed to concrete, which may facilitate external attack and cracks formation
form ettringite. The concentration of SO24 ions needed for shot- [134]. According to the authors, dry-mixed shotcrete (with fly ash,
crete to be damaged by thaumasite sulfate attack is dependent low alkali accelerator and water-reducing agent) performs much
on the Ca/Si ratio in the C-S-H and the presence/absence of worse than ordinary concrete when exposed to the combined
portlandite [123]: the lower the calcium content, the higher the effect of sulfate attack (10% Na2SO4 immersion and 60 °C drying)
SO2
4 concentration required. However, field studies have shown and drying-wetting cycles. Details about the composition of the
that highly concentrated pore solutions and low H2O activities ordinary concrete are not provided and so, direct comparison of
are associated with gypsum, bassanite (CaSO40.5H2O) and thaum- the different mixes and their properties is not valid. Resulting dam-
asite formation [16]. Such pore solutions can be linked to drying- age may be attributed to poor shooting of the shotcrete.
wetting cycles, which are known to speed up deleterious processes The opposite has also been reported: shotcrete performed bet-
in concrete [124] and shotcrete [125]. ter (when exposed to 10% Na2SO4 immersion and cycles of 60 °C
Other less common types of sulfate attack include efflores- drying) than concrete with a similar composition [133,135]. The
cences: sulfate salts, like thenardite (Na2SO4) and mirabilite authors attributed the better properties to the microstructure, i.e.
(Na2SO410H2O) precipitate in pores, leading eventually to further the numerous micropores in the shotcrete could accommodate
expansion and physical damage [126–128]. degradation products like ettringite and gypsum. In steel fibre rein-
forced shotcrete the fibres can eliminate the swelling pressure
4.2. Role of accelerators on sulfate attack from secondary ettringite and gypsum, performing even better
[133]. The damage caused by sulfate attack on shotcrete exposed
All aspects considered in the previous section (4.1.) apply to the same cycles (drying at 60 °C for 8 h followed by immersion
equally to shotcrete and concrete; however, in the case of shotcrete in 10% Na2SO4 solution for 16 h), characterized by ultrasound prop-
the important contribution of the admixtures needs to be consid- agation methods, was classified by Wang et al. [136] in 3 stages:
ered, especially alkali-free accelerators containing aluminate and ettringite action, ettringite + gypsum action and gypsum action.
sulfate solutions. The influence of the admixtures added to shot- Similar to what happens for concrete, the Na/Mg-SO4 -type
crete on the resistance to sulfate attack has been studied, but it solutions used in the standards are not representative for typical
is still a hot topic. Alkali-free accelerators are nowadays preferred groundwater or drainage solutions in contact with shotcrete. Cal-
to alkaline ones because of the long-term strength development, cium sulfate-type solutions are the most commonly found ones;
health and safety considerations, and their effect on decreasing however, the corresponding tests are much slower and not suitable
leachability and reducing sintering in the drainage system. How- for accelerated lab procedures. Since the concentration and the
ever, this type of accelerators provides aluminate and sulfate into type of sulfate used are not realistic, the results from the current
the cement and thus its dosage (up to 12 wt% with respect to bin- lab tests do not give enough insight on the mechanism of the
der mass) must be controlled and counterbalanced by the amount attack. Lorente et al. [137] compared natural and accelerated
and type of cement used. Several authors have compared the use of methods with Na- and Mg-sulfate solutions obtaining ‘contradic-
alkaline and alkali-free accelerators in terms of the response of the tory’ results: in the migration tests specimens exposed to Mg2+
corresponding shotcrete to sulfate attack. According to Paglia et al. performed worse (showed larger penetration depth) than those
[129], the sulfate resistance of shotcrete mixes is mainly controlled exposed to Na+; in the diffusion experiments, the opposite hap-
by the concentration of Al3+ and SO2 4 from the admixtures and the pened, the explanation relying probably on the kinetics of brucite
portlandite availability, affecting both the decalcification rate of C- formation. This example, obtained for concrete but applicable for
S-H and its disintegration. Paglia et al. [130] found a correlation shotcrete, illustrates the inadequacy of accelerated tests methods.
between length changes in the specimens and the total SO2 4 As for now, there is still no widely accepted and suitable test
adsorbed, and the SO24-OH exchange (in 5% Na2SO4 immersion method mimicking all different aspects of sulfate attack [138].
tests). Additionally, they reported a correlation between long-
term expansion of shotcrete and short-term expansion of (lab) 4.4. Real damage reports in shotcrete
cement paste specimens.
Lee et al. [131] used a 7% Ca(NO3)2 aqueous solution as alkali- Thaumasite sulfate attack in shotcrete has been mainly covered
free accelerator and compared the behaviour of samples immersed in the literature by field studies [16,139–147]. Mostly, thaumasite
in Na2SO4 solutions (refreshed every 4 weeks up to 1 year and then formation occurs via coupled dissolution of portlandite and car-
every 3 months). Mortars made with accelerator performed much bonate fillers/aggregates, decalcification of C-S-H and transforma-
worse than those without and, as expected, increasing SO2 4 con- tion of pre-existing ettringite (substitution of Al by Si and
centrations increased the damage. The explanation lies on the SO3 + H2O by CO3), given that external sources of aqueous SO2 4
higher permeability, lower strength and thus easier ingress of are available [148]. In some cases, secondary calcite and brucite
water and SO2 4 ions. Ettringite formation is promoted by the addi- form together with thaumasite (Fig. 6f). In this line, Mittermayr
tion of the alkali-free admixture. The authors reported no contribu- et al. [16] studied the formation of thaumasite in several tunnels
tion of gypsum to the degradation, mainly due to ettringite and in Austria. Mineralogical, chemical and isotope methods were used
thaumasite formation. The use of aluminium-based accelerating to characterise the damaged shotcrete as well as the groundwater,
admixtures has also been reported to favour thaumasite formation drainage water and pore solutions. The authors found evidence of
[132], being the effect higher with aluminium sulfate containing evaporative enrichment of especially Na+, Cl and SO2 4 ions in the
admixtures than with sodium aluminate ones. pore solutions within wetting–drying cycles, but no presence of
ettringite, believed to have previously dissolved at a specific pH
4.3. Other parameters controlling the sulfate resistance of shotcrete range (e.g. below pH 9). To determine the SO2 4 ion source required
to form thaumasite, Mittermayr et al. used the stable isotopes of
Most reported studies use the ASTM C1012 standard sulfur (i.e. ratio of 34S/32S) [148]: oxidative weathering of pyrite
(5% Na2SO4) [270], the Chinese Standard GB/T 50082-200960 and dissolution of marine evaporates such as gypsum and
(10% Na2SO4) [133], or similar tests with Na2SO4 solutions ranging anhydrite (CaSO4) were found to be the actual SO2 4 ion source
478 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

for sulfate attack, whereas no contribution was attributed to atmo- has reached its final stage (total destruction), only the most stable
spheric SO2, internal sulfate in the cement or combustion soot. phases under the current environmental conditions remain.
Four different case studies of thaumasite formation with car- Thus, chemical sulfate resistance alone is not the key to durable
bonation in Norwegian concrete and shotcrete are covered by shotcrete facing sulfate attack. In order to recommend proper mea-
Hagelia et al. and Hagelia [140,141]. In Norway, sprayed concretes sures for sulfate resistant shotcrete the underlying succession of
with sulfate resistant OPC are often employed in direct contact reactions have to be taken into account. Important initial stage
with sulfide bearing rock types such as the Alum Shale Formation mechanism might be cracking due to rock mass loads and shrink-
(containing calcite, pyrrhotite, etc.). Thaumasite formation was fre- age, non-watertight constructive elements such as joints, deficits
quently associated with contemporaneous and late stage decalcifi- or permeable (low quality) shotcrete [16]. Consequently, capillary
cation of C-S-H, which also led to the formation of internal suction and/or wick action are responsible for the transport of
secondary calcite (popcorn calcite like). In some cases, steel fibres groundwater through the shotcrete. In contact with the tunnel
were corroded due to internal carbonation, which had also resulted atmosphere the typical evaporation and drying-wetting phe-
in a porous matrix. The thaumasite formation-carbonation process nomenon can occur stimulating both chemical and physical deteri-
caused a steady drop in the pH, from 13 to neutral, and generally oration as described above. The following recommendations can be
took place in wet and cool environments, thaumasite being stable given:
at pH 6 to 11 and <15 °C [149,150]. The reaction progress was Formation of cracks should be avoided and the use of construc-
found to be sensitive to layer thickness; however, the attack was tive elements such as unsealed joints should be prevented. The
not very sensitive to the C3A content (within 1–8%) or to the dis- transport of groundwater (especially very slow flowing or drip-
persion degree of the silica fume. ping) through the shotcrete lining should be diminished in order
40 years old tunnels in China, drained by sulfate-containing to avoid evaporation, followed by extremely concentrated and
groundwater, and showing intense damage, were studied by min- harmful solutions. Low porosity shotcrete and proper curing are
eralogical, chemical and imaging techniques [142]: gypsum, ettrin- required to further reduce the risk of capillary action.
gite and thaumasite were found. Two sources of SO2 4 ions were In general, the sulfate resistance of shotcrete is determined by
identified, i.e. groundwater and oxidized coal powder originally (i) the C3A content in the cement and the Al2O3 content in the
containing sulfur (from the trains loading). Temperature and rela- SCMs, (ii) the water-soluble Al2O3 in the accelerator, and (iii) the
tive humidity conditions in the tunnels were found to be ideal for impermeability of the shotcrete structure [151]. However, it is
thaumasite sulfate attack. However, the damage was also attribu- not recommended to rely on the chemical resistance of C3A-free
ted to the poor quality of the concrete and shotcrete installed: thin cements (SR0 according to EN 197-1 [271]) as studies have shown
linings, low strength, high permeability and no waterproofing that damage related to thaumasite sulfate attack can still occur
measures. [121,140,152]. It seems to be more important that a low water/bin-
Parts of the shotcrete lining of a tunnel in Germany exposed to der ratio (0.45) is used. Further improved physical and chemical
steam locomotive soot started to break off from its stonework sup- robustness can be achieved by implementing suitable active SCMs
port after 40 years in service [143]. Ettringite was found on the rear such as flay ash, granulated blast furnace slag and silica fume. The
surface of the shotcrete together with soot. Thaumasite had formed usage of active SCMs leads to lower amounts of portlandite, and
at the concrete interface and up to 15 mm inside the shotcrete, fill- permeability is in most cases lower [153]. Studies from Hagelia
ing microcracks and veins parallel to the shotcrete surface, leading et al. and Hagelia [140,141] report that silica fume delayed forma-
to great expansion of the shotcrete. The deterioration found in a tion of thaumasite in shotcrete but did not inhibit it completely.
shotcrete tunnel in Dallas was investigated by Puppala et al. Mittermayr et al. [154] showed that a shotcrete with CEM I SR0,
[146]. The SO24 content was measured as BaSO4 after reaction with 30 wt% of blast furnace slag and 3 wt% of silica fume, exhibited
BaCl2 (maintaining pH at 5–7). Qualitative XRD revealed the pres- excellent resistance to thaumasite sulfate attack in comparison to
ence of ettringite and thaumasite. The authors encountered SO2 4 conventional sulfate resistant shotcrete based on CEM I SR0 only.
levels of up to 2000 ppm in affected areas and a moisture content If ‘‘spray cement” is used (to produce dry-mix shotcrete with no
of 14%. accelerator) in sulfate-loaded environments, restrictions on the
Thaumasite was also found in Swiss tunnels [144], both in the SO3, C3A and Al2O3 content in the cement apply. In the presence
leached zone and in sound cement paste zones. According to the of water with SO2 4 above 600 mg/l, the SO3 content has to
authors, low temperature does not seem to be required for thaum- be 3.5 wt%; in all cases, the C3A and Al2O3 content must be 1.0
asite formation in tunnels: in some of the cases the temperature and 5.0 wt%, respectively [9]. Restrictions on the SO3 and Al2O3
was even higher than 20 °C. The severity of the deterioration content also apply to the setting accelerators: according to the Aus-
depended mainly on the chemical composition and amount of trian Building Technology Association (ÖBV) Guideline, the Al2O3
groundwater discharge, the thickness of the shotcrete and concrete wt.% times the accelerator dosage (in % of binder) must
layers, porosity and permeability of the linings, and the flowing be 115 wt%; the sum of the SO3 from the cement and the acceler-
pathways through the shotcrete and concrete. The concentration ator should be kept 4.8 wt%. In shotcrete exposed to sulfate envi-
of SO2
4 ions in the water and the use of low Al cements did not ronments (200<SO2 4 <600 mg/l) the addition of mineral fillers is
seem to be very important. Romer et al. [144] stated that shotcrete only allowed if the CaCO3 content is kept below 10%. In environ-
is more prone to thaumasite formation because of the inhomoge- ments where the SO2 4 1500 mg/l, the use of shotcrete as a perma-
neous distribution of porosity. nent support structure is not permitted [9].

4.5. Recommendations and countermeasures against sulfate attack on

shotcrete 5. Freeze-thaw

As described in the previously referred field studies, sulfate 5.1. Mechanisms and effects of freeze-thaw and controlling parameters
attack in tunnels is never a stand-alone attack. It is rather a com-
bined form of attacks that may ultimately lead to the total destruc- Cyclic freezing and melting as well as deicing salt exposure can
tion of the shotcrete. For instance, in some cases the destroyed significantly affect the durability of both wet- and dry-mix shot-
shotcrete is reported to consist only of a mush of thaumasite, cal- crete. The resultant shotcrete damage is attributable mainly to
cite and aggregates. The reason for this is, that once deterioration physical attack, related to the expansion of freezing water in con-
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 479

nected capillary pores, leading to (i) internal damage, with reduc- The quality of the mortar matrix and the coarse aggregates
tion in Young modulus, (ii) stresses and cracks formation, (iii) sur- determine the frost resistance of the shotcrete; if sound aggregates
face scaling (surface damage due to progressive removal of thin are used, ‘failure under frost occurs by the weakening and break-
flakes), especially when in contact with salts, and (iv) loss of adhe- down of the cement mortar which binds the coarse aggregate
sion between shotcrete and rock (Fig. 7). together’ [161]. More specifically, the resistance depends on the
Several mechanisms have been proposed to explain degradation contents of cement, water and air, and on the grading and mix
of concrete/shotcrete exposed to frost: (i) volume expansion of design, having the type of cement used little effect. Indeed, two
water during freezing in confined pores, (ii) growth of ice crystals, types of ‘air’ have to be distinguished: (i) ordinary entrapped air,
(iii) moisture movement, (iv) hydraulic pressure caused by the which is not favourable for freeze-thaw resistance, and (ii) small
movement of water through capillaries, (v) expansion due to freez- entrained air bubbles, which interrupt the capillary pore system,
ing in the surface pores forcing water inwards through small pores therefore being very beneficial to prevent frost attack. In the case
[23,155,156]. According to Chatterji [157,158], ice infiltrates por- of shotcrete, many of the larger air bubbles are lost during
ous materials in the form of ‘dendrites’ (i.e. branched, tree-like shooting.
crystal extensions), encapsulating in the process part of the still Common indicators of freeze-thaw damage in shotcrete/con-
unfrozen water. During further cooling, the unfrozen water devel- crete are the reduction of the dynamic elastic modulus and weight
ops high hydrostatic pressure capable of causing damage. Based on losses. For a shotcrete to qualify as ‘‘frost-resistant”, according to
experimental measurements of expansion, acoustic emissions and the ÖBV Guideline [9], the reduction of the static modulus of elas-
calorimetry on various types of concrete, Kaufmann [159] pro- ticity must not exceed 25% after 56 freeze-thaw cycles. To assess
posed a model to describe the freeze-thaw process of a liquid in the freeze-thaw resistance of concrete the ‘critical degree of satu-
a pore system divided in five phases: (i) freezing initiation (0 to ration method’ was proposed by Fagerlund [162]. The method is
5 °C), (ii) penetration of the ice front (0 to 10 °C), (iii) redistribu- based on the assumption that there are critical moisture contents
tion of pore solution (-5 °C to minimum temperature of the frost or degrees of saturation at freezing. As such, the difference
cycle), (iv) thawing (minimum temperature to 0 °C), and (v) resat- between the critical degree of saturation SCR (determined by sub-
uration (above 0 °C). All five phases contribute to the deterioration, jecting sealed specimens containing different amounts of water
the importance of each phase being dependant on the pore size to freeze-thaw cycles) and the capillary degree of saturation SCAP
distribution and on the climatic conditions (i.e. relative humidity (measured as the ability of specimens to absorb water as a function
and temperature). Coussy and Monteiro [160] used poromechanics of time) gives a ‘potential freeze-thaw resistance’ of a certain con-
(i.e. study of the mechanical behaviour of porous media) to study crete type used in different environments. The freeze-thaw resis-
the stresses and strains that develop as a result of freeze-thaw pro- tance is expressed as F = SCR - SCAP. If the SCAP exceeds the SCR
cesses, including hydraulic pressure development and cryo- after a short time of water exposure, the concrete is considered
suction. Their analysis indicates that ‘the air voids act both as to have a poor freeze/thaw resistance. This means the longer F
expansion reservoirs and efficient cryo-pumps’, thus reducing the has a positive value the better the quality of the concrete versus
pressure built up during freezing. freeze/thaw exposure.

Fig. 7. Shotcrete damaged from freeze-thaw exposure (a, b). Scheme of one possible model for freeze-thaw damage progression: (i) shrinkage and cracking due to cement
phase dehydration and ice invasion in capillary/gel pores (c), and (ii) pressure release during deicing, external water inflow and gel expansion in water-saturated
micro-/macrocracks (d).
480 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

Better behaviour in terms of freeze-thaw resistance has been resistance to freeze-thaw [169,170], follows a protocol of various
reported for shotcrete versus concrete without accelerator [135]. steps of saturation under water, drying, and pressure saturation,
The explanation probably lies on the numerous micropores that coupled with measurements of moisture content, maximum suc-
developed in the matrix of the shotcrete due to the shooting with tion absorption, volume and density, dry weight and
compressed air, which ‘helped relieve the frost-heaving pressure macroporosity.
on the capillary pore water and the osmotic pressure of the super- Acoustic emissions and temperature measurements have been
cooled water migration’. The formation of large amounts of ettrin- used to monitor initiation and development of microcracks in the
gite during the first minutes of hydration in aluminium sulfate- shotcrete-rock interface due to freeze-thaw exposure [171]. The
based accelerated systems has also been proposed as a hypothesis number of acoustic events were recorded, together with location,
to explain the better behaviour of shotcrete versus concrete: for- magnitude and seismic energy, and then interpreted and related
mation of ettringite with the consequent ‘trapping’ of water (32 to what is happening: debonding, cracks formation and growth
molecules of water per unit formula) reduces the amount of free (also probably due to ice formation and melting). This technique
water available to freeze [163]. Moreover, the composition of the has recently been also applied to monitor freezing damage in
water should play an important role: higher concentration of ions alpine rock walls. Advanced techniques, like computed tomogra-
in the free water of accelerated systems would also decrease the phy, have also been proposed to study freeze-thaw damage in
freezing point. shotcrete [172]: appearance of micropores and fractures on the
On the other hand, set accelerators may also reduce the resis- cemented surface of the aggregates and at the corners of the spec-
tance to deicer salt scaling and increase the shrinkage. According imens can be monitored.
to Jolin et al. [164], carbonate-based accelerators, which affect both
initial and final set, are worse than aluminate-based ones, which 5.3. Recommendations to reduce freeze-thaw damage
only affect the initial set and ‘have probably a reduced impact on
the development of the microstructure’. According to Park et al. The use of air-entraining agents is a common practice to
[165], the use of amorphous dodecacalcium hepta-aluminate improve the freeze-thaw resistance (and also workability) of con-
(C12A7) mineral-based accelerator leads to better behaviour in crete and shotcrete [164,173–177]. However, air-bubble entrain-
freeze-thaw tests than aluminate and silicate-based accelerators, ment in fresh concrete and shotcrete is a complex problem. On
and almost to the same results as ‘plain shotcrete’. This perfor- the one side, every component in the mix, as well as the mixing
mance is attributed to the lower permeability of the mixes made process, batch quantities and placing methods, affect the air void
with C12A7, which resulted in denser microstructure due to the system; on the other side, it is very difficult to routinely character-
large amounts of ettringite formed during the initial stages of ize air void systems in fresh concrete and shotcrete directly in the
hydration. In the same line, according to Won et al. [166], shotcrete field [178]. The most commonly used indicators to characterize the
mixes with high strength C12A7 accelerator containing calcium sul- entrained air are the spacing factor (i.e. maximum distance
foaluminate and granulated blast furnace slag, perform better after between air voids in the cement paste), the total volume of air
freeze-thaw cycles than mixes with standard C12A7, which only voids, the specific surface of the air voids, and the A300 (content
contained the aluminate phase, because of the reduction of micro- of micropores <0.3 mm), as described in the standard EN 480-11
voids and the suppression of water penetration. For similar reasons, [275].
high strength shotcrete (with silica fume, lower water/cement (w/ In the case of shotcrete, most of the air is removed due to the
c) ratio and 8% alkali-free aluminium sulfate-based accelerator) shooting process; however, the small air voids that remain in the
performs better in terms of surface scaling than a shotcrete with matrix contribute to improve the behavior towards frost and
5% of alkaline aluminate-based accelerator, no silica fume and freeze-thaw exposure. Lukas and Kusterle [176], Kusterle [177]
higher w/c ratio, when exposed to accelerated freeze-thaw tests and Jolin et al. [164] reported on the use of air-entraining admix-
consisting of 300 temperature cycles from 4 to 18 °C [167]. tures in dry-mix shotcrete, confirming the positive effect towards
frost and salt scaling resistance. Chen et al. [174] also measured
5.2. Freeze-thaw in tunnels and testing methods improvement of freeze-thaw resistance using air-entraining agents
in wet-mixes: lower mass losses, smaller reduction of relative
Shotcrete in tunnels can be prone to freeze-thaw damage dynamic modulus, larger relative ultrasonic wave velocities, and
because of the exposure to water leaking from the rock and the lower loss of compressive strength. Both liquid and powdered
low temperatures reached in many tunnelling structures in the air-entraining admixtures are currently used. Disadvantages on
winter months [168]. However, deterioration from freeze-thaw is the use of liquid products for dry-mix shotcrete are related to
nowadays prevented in many tunnels by the use of double shell the difficulty to control the amount added to the shooting water
lining in combination with deicing salts. In single shell tunnels and the variations due to the nozzle man changing the shooting
though, freeze-thaw can be a problem, especially in the regions consistency [179]. Powdered and micro-sphere-based products
near the portal, where the temperature differences in the freeze- are easier to dose and help producing a more homogeneous air
thaw cycles are greater. According to Holter et al. [169], the mois- void system.
ture content in tunnel linings is likely to be below a critical degree Air entrainment is generally accompanied by the reduction in
of saturation for freezing damage and so, sprayed concrete lining strength. However, strength reduction due to the increase of air
structures waterproofed with membrane are not prone to freeze- pores can be overcome by adding silica fume: Choi et al. [175]
thaw deterioration. According to the authors, the conditions in tun- reported a satisfactory spacing factor and a good retention of small
nels are not well represented by the standard tests CEN/TS 12390– entrained air bubbles (which improved freeze-thaw and scaling
9 [272] and ASTM C666 [273], which reckon much more severe resistance) in wet-mix shotcrete samples with air-entraining
conditions than the worst possible case in tunnels. They propose agents and silica fume replacement.
a more realistic method (called ‘pore protection factor method’), Reinforcing the shotcrete with steel fibres is another method
where the critical freezing dilation is tested according to ASTM to improve its frost resistance [168]: the fibres help controlling
C671 [274] (with no pre-drying requirements before freezing), cracks development and propagation by sharing the frost heaving
rather than by the critical degree of saturation according to pressure and stress concentration during exposure to frost condi-
Fagerlund [162]. This method, developed in Finland and Norway tions. The use of polypropylene fibres (PP) in shotcrete has been
to calculate porosity and moisture content profiles, and thus assess and still is a research topic. Kan et al. [180] found an optimum
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 481

volume of PP fibres to improve frost resistance at 0,9% (of cemen- is counterbalanced by the reduction of another substance (e.g. O2
titious material). Morgan et al. [181] reported excellent frost or H+) at the metallic surface (Eqs. (18–22)).
resistance of a shotcrete prepared with PP fibres and high volume
of fly ash (60 and 75% mass of total cementitious material): an FeoðsÞ ! Fe2þ þ 2e ð18Þ
optimum was found for 400–450 kg/m3 cementitious material
and 4–6 kg/m3 PP fibres. Similarly, shotcrete with PP fibres 2Fe2þ ! 2Fe3þ þ 2e ð19Þ
(3–4.5 kg/m3), silica fume (5–8% mass of total cementitious mate-
rial) and high volume fly ash (60% mass of total cementitious 0:5O2ðaqÞ þ H2 O þ 2e ! 2OH ð20Þ
material) performed satisfactory after freeze-thaw exposure,
according to Zhang et al. [182]. 2Hþ þ 2e ! H2ðgÞ ð21Þ
Other ways to improve freeze-thaw resistance, like increasing
the proportion of coarse aggregates and the fineness modulus of 2Fe3þ þ O2ðaqÞ þ 2OH ! 2FeOOHðsÞ ð22Þ
the fine aggregates, have been reported [183]; however, the under-
lying mechanisms are not well understood yet. The inclusion of The resistance to chloride and carbonation front penetration
other additives used to improve frost resistance of concrete, like depends, among others, on the composition of the binder and on
synthetic zeolites, organic resins, encapsulated siloxane, fibres the permeability of the shotcrete, in turn controlled by the cement
and nanoparticles [184], remains an open field for shotcrete type, water and moisture content, the curing, etc. Humidity plays a
research. key role on both types of attack: saturated pores make ion diffu-
According to EFNARC [1], if the sprayed concrete will be sion difficult while in very dry pores dissolution is inhibited. Apart
exposed to freeze-thaw in moderate water saturation without salt from the effects on the steel reinforcement, chloride ingress
(EN 206 class) no further requirements for testing are requested. If increases the solubilisation of the cement phases [189] at a local
exposed to more aggressive environments, specified in EN 206, scale, whereas carbonation leads to local volume and strength
then it shall fulfil the requirements as in the standards for scaling changes at the active carbonation front. In practice, steel corrosion
and freeze-thaw resistance. To test freeze-thaw resistance, EFNARC can be recognized by the formation of cracks, and the accumula-
recommends the ASTM C666 [273], and for scaling resistance the tion of rust and rust-derived products, both at the steel-shotcrete
ASTM C672 [276] or the Swedish standard SS 137244 [277]. For interface and on the surface of the shotcrete.
the latter, samples obtained from drill cores or cast cubes, are Chloride ions from the ‘outer’ environment (i.e. from external
sealed in rubber cloth leaving one surface exposed for testing. They sources, like deicing salts, sea water, effluents, etc.) can penetrate
are then covered by the test solution (3% NaCl solution or water) to the pore structure of shotcrete by diffusion and capillary forces.
a depth of 3 mm and submitted to a 24 h temperature cycle, rang- Some of the chloride react with the cement paste to form Friedel’s
ing from 16 to 24 °C to 14 – 20 °C. After 7, 14, 28, 42 and 56 salt (chloride-AFm) and is thus not available for further penetra-
(and sometimes up to 112) cycles, the scaled material is weighed tion [190]. The Cl that remains in the pore solution can then pro-
and the frost resistance categorised, according to the weight loss gress towards the reinforcement. Once Cl reaches the steel, the
(e.g. after 56 cycles: <0.1 kg/m2 very good; <0.5 kg/m2 good; passive layer is rapidly de-stabilised by the small pits created by
and <1.0 kg/m2 acceptable) [2]. Another standard for freeze-thaw the Cl, which act as anodes. Immediately at the metal/Cl/oxide
resistance tests reported in the literature is the Chinese Standard interface the iron dissolves, and then reacts to form the solid cor-
GB/T 50082-2009 [185]. It involves 24 days of curing, soaking in rosion product rust (e.g. FeOOH) and hydrogen gas (H2) (Eqs.
water for 4 days and freeze-thaw cycles. The dynamic modulus (18–22)). The pits and cracks formed on the surface of the steel
of elasticity, weight losses, ultrasonic waves propagation are mea- reinforcement progress inwards, leading to local breakage of the
sured, and the pore structure (i.e. bubble diameter and spacing fac- rebar.
tors) are analysed by a pore structure tester [174]. The Austrian In the case of carbonation the reaction between dissolved basic
Guideline for shotcrete stipulates that frost-resistant shotcrete components of the cement paste and atmospheric CO2 leads to a
should not show a reduction of the static modulus of elasticity decrease of the pH values of 12–13 to below 8 (Eqs. (9–12)). The
(measured according to ÖNORM B 3303 [269]) above 25% after carbonation front progresses inwards until it reaches the reinforce-
56 freeze-thaw cycles [9]. In the presence of deicing salts, the stan- ment, shifting then the steel regime from passivation to corrosion.
dards for concrete ([278] and [279]) should also be used for
sprayed concrete. Two tests that were developed to assess the 6.2. Cl penetration and carbonation in shotcrete
resistance of concrete to freeze-thaw are also used for shotcrete:
the CDF-test (capillary suction, deicing agent and freeze-thaw test) Carbonation and chloride penetration induced corrosion are not
and the CIF test (capillary suction, internal damage and freeze- considered to be particularly dangerous for subsurface double shell
thaw test) [186–188]. shotcrete structures, whether mesh or fibre reinforced, especially
when compared with other durability issues, like sulfate attack.
However, it is an important issue in concrete repair and in single
6. Electrochemical corrosion shell tunnels (Fig. 8). As all these phenomena are related to perme-
ability and ‘reactivity’ of the shotcrete and concrete layer(s), char-
6.1. Processes underlying electrochemical corrosion acteristics like the chloride diffusion coefficient or the carbonation
rate are commonly used to classify and compare the quality of
Steel embedded in shotcrete is physically protected by the shot- shotcrete and concrete mixes.
crete itself, and chemically by the high pH of the pore solution and One of the main differences between shotcrete, especially in the
the passive layer that develops under those conditions on the steel case of double shell tunnels, and concrete is the fact that shotcrete
surface. However, carbonation and chloride penetration may is covered by concrete, preventing access from atmospheric CO2
destabilize the passive layer, leading to steel corrosion (Fig. 8a- and/or Cl ions from the ‘external’ environment. At the same time,
d). The dominant mechanism underlying steel corrosion in shot- shotcrete is directly exposed to the rock and related percolating
crete and concrete is of an electrochemical nature. A pre- waters, which contain a wide range of concentrations of dissolved
requisite for this process is the establishment of an electrochemical Cl and inorganic species. Interaction between shotcrete and per-
potential difference, where oxidation of the metal alloy (e.g. steel) colating water, in turn, contributes to leaching and subsequent
482 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

Measured
carbonation front
Steel
depassivation

CO2(g) Ca2+ C-S-H

CH

pH 8-10 pH 12-13

Fig. 8. Examples of steel corrosion in reinforced shotcrete structures (a, b, c). Scheme of carbonation and steel corrosion (d). Note the carbonation front represents the
phenolphthalein measured front. Once the steel is depassivated due to the lower pH of the carbonated concrete, the steel bar may start corroding when enough humidity and
oxygen concentration are present.

massive formation of CaCO3 in the drainage system of tunnels (Sec- conventional shotcrete in terms of carbonation [167]. Considering
tions 2 and 3). that the high performance shotcrete contained silica fume, had a
Specific cases where Cl penetration in shotcrete is relevant are: lower w/c ratio and 8% of alkali-free accelerator, in contrast to
subsea applications, exposure to Cl containing deicing salts or in the 5% of alkaline aluminate accelerator of the conventional shot-
structures close to the seashore. In such aggressive environments, crete, with no silica fume and higher w/c ratio, lower permeability
the thickness of the shotcrete layer versus the service life time and higher resistance to Cl and CO2 penetration are expected.
expectation is a crucial parameter. In Norway, the minimum thick- The type and dosage of accelerator determine the aluminium/-
ness of sprayed concrete layers was increased from 2-5 cm to 6 cm sulfate ratio in the system, which in turn controls the potential for
from 1997 to 2010 to prevent carbonation, electrochemical corro- Friedel’s salt formation, and thus the capability of the shotcrete to
sion and salt water attack. Back in 2011, to increase the life time retard chloride penetration. The influence of the type of accelerator
of sprayed concrete from 50 to 100 years the thickness was then on the resistance to Cl and CO2 penetration has been studied by a
specified as 8 cm [191]. In subsea applications, the resulting dam- few authors. According to Park et al. [165], the use of amorphous
age of shotcrete structures is very often linked to multiple com- C12A7 mineral-based accelerators for wet-mixed shotcrete leads
bined reactions, including microorganisms’ contribution and to lower Cl penetration than other commonly used silicate-
alkali-aggregate reaction (Section 7). The use of deicing salts, espe- based or aluminate-based accelerators, but to almost the same val-
cially those based on calcium chloride can also lead to the forma- ues as for plain concrete without accelerator. The higher resistance
tion of calcium hydroxychloride hydrates (xCa(OH)2yCaCl2zH2O), is explained by the development of a microscale ‘web’ structure,
and subsequent expansion and cracking in hardened shotcrete/con- associated with the formation of ettringite, within the first hour.
crete [192]. High strength C12A7 accelerator is also reported to produce more
Most of the literature reporting on chloride and carbonation resistant matrices towards Cl and CO2 penetration than the com-
resistance of shotcrete provides values of measured ion penetra- monly used C12A7 [166].
tion, as from standard tests, and compares different samples. Wang
et al. [193] reported worse properties for ordinary shotcrete in 6.3. Recommendations for reducing electrochemical corrosion of
terms of Cl penetration compared to ordinary concrete, which shotcrete
the authors attributed to the porosity generated during shooting
and to the ‘looseness’ of the microstructure due to the fast forma- Steel reinforcements in shotcrete can be placed in the form of
tion of ettringite and C-S-H gel. The same authors reported lower rebars or fibres, whereby its use typically depends on country pref-
carbonation depths in shotcrete than in concrete when exposed erences, guidelines, rock type and the aggressiveness of the envi-
to 20% CO2, 70% relative humidity and 20 °C, which they attributed ronment. Under conditions where conventional reinforcement
to the lower density of the shotcrete matrix. These contradictory presents a high corrosion rate, steel fibres may be unaffected
results are not supported by microstructural and/or compositional [194]. In areas where the rock is predominantly composed of gran-
data. High performance shotcrete was found to behave better than ite (Scandinavia) the use of steel arches to reinforce the shotcrete is
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 483

not needed and fibre-reinforced shotcrete is preferred; in regions available so far, but it appears to be quite frequently coupled with
with inhomogeneous and/or not well consolidated rocks (Alps) other attacks in some shotcrete applications. Apart from these
the use of steel mesh is often more effective to provide the extra issues, it is important to mention that the deformations of the rock,
support required. the blasting, etc., may lead to mechanical overloading, cracking,
The length (1–6 cm), composition and distribution of the shear failure and creep failure, and thus condition the properties
fibres in the shotcrete matrix determine the corrosion resistance of the shotcrete and its durability. In special cases, compressed
as well as the mechanical properties of the shotcrete. In aggressive air is used to compensate the pressure from the water ingress into
environments, where steel fibres in reinforced shotcrete corrode, the tunnel [206–208]. This air may penetrate the young shotcrete
the use of synthetic (e.g. PP) fibres is proposed as a means to pre- causing local porosity increase and providing extra pathways for
vent electrochemical corrosion [195–197], this being especially chemical attacks.
relevant for cracked shotcrete [198]. PP fibres have, however, been
banned in some countries due to their pollution potential, espe-
7.1. Alkali-aggregate reaction
cially in undersea tunnels, where the fibres from the rebound
material remain floating in the ocean [199].
Alkali-aggregate reaction takes place over time between the
As already mentioned, proper application of the shotcrete to the
highly alkaline cement paste and reactive aggregates. This reaction
substrate is crucial to produce homogeneous matrices with a low
typically causes expansion, cracking and loss of strength of the
permeability [200]. Similarly to conventional concrete, the addi-
altered shotcrete and concrete structures (Fig. 9a). Three different
tion of silica fume and granulated blast furnace slag may improve
forms of alkali-aggregate reactions can be distinguished: (i) alkali-
the resistance to Cl penetration due to the reduction of total per-
silica, (ii) alkali-silicate and (iii) alkali-carbonate, with alkali-silica
meability [201]. Additions of SCMs to shotcrete, however, may
being by far the most important reaction. Generally, alkali-silica
have the opposite effect on the carbonation resistance: decreasing
reaction takes place between alkali hydroxides (i.e. NaOH and
the portlandite content in the paste also reduces its ability to pre-
KOH) from the cement paste and related pore solutions and reac-
vent further ingress of the carbonation front towards the steel. This
tive, poorly crystalline silica containing aggregates. For the reac-
means, when low portlandite binders are selected to reduce the
tion to take place, apart from sufficient concentrations of alkalis
calcium leaching and the sintering potential (Section 3), greater
and reactive silica/silicic acid, high relative humidity and presence
thicknesses of the shotcrete layers will be needed to ensure that
of water are needed. The reaction leads to the formation of a silica-
no electrochemical corrosion due to carbonation will take place
based gel in the cracks, around the aggregates and also at the sur-
during the service life of the shotcrete.
face of the damaged shotcrete and concrete [23]. However, K+ and
The impact of nanoparticles (SiO2, CaCO3 and Al2O3) on poros-
Na+ ions do not take part in the early stages of this reaction as they
ity, pore size distribution and Cl permeability (and strength) of
are very soluble, and thus remain in the pore solution. Indeed, pro-
shotcrete, prepared with liquid aluminate-based accelerators and
gressive attack of OH– ions breaks the Si-O-Si bonds and loosens
superplasticizers, was analysed by Li et al. [202]. According to
the structural network of silica-bearing phases, producing nega-
the authors, the additions of ultrafine SiO2 (4%) improved both
tively charged deprotonated monosilicic acid, sodium silicate aquo
strength and resistance to Cl penetration, and nanoparticulate
complexes and a huge variety of negatively charged polysilicic acid
CaCO3 (0.5%) improved even more the resistance to Cl ingress
(i.e. pairs of Si-O groups) to which then the positively charged
but not the strength. The use of corrosion inhibitors established
alkalis get attached and an alkali-silica gel (N/K-S-H) forms (Eq.
for reinforced concrete has not been reported for shotcrete
(23)). The hygroscopic gel can absorb water, leading to swelling
structures.
and further crack formation at advanced stages. The availability
The aluminium/sulfate ratio in the shotcrete, to which both
of Ca2+ ions also plays a role in the process: during the reaction
cement and accelerator contribute, could be optimized to produce
of the alkalis and OH– with the silica, available Ca2+ ions are also
a higher Friedel’s salt content; however, this ratio has to be firstly
adsorbed on the silica gel and C-S-H may precipitate [209,210].
optimized for ettringite formation, setting time, early strength
The modelling of the process was reviewed by Pan et al. [211].
development, etc., relegating Friedel’s salt formation potential to
Accordingly, if the equivalent Na2O content in concrete is kept
a secondary subject.
below 3–4 kg/m3, the risk for alkali-silica reaction is very low.
The standards used and reported in the literature for carbona-
tion and Cl penetration of shotcrete include the ASTM C1202 
Naþ þ H3 SiO4 þ H2 O ! N  S  H ð23Þ
Cl penetration test [203], NT Build 492-1999.11 rapid Cl migra-
tion test [204], JTG/T B07-01-2006 Specification for deterioration Alkali-silicate reaction (which involves the attack of sheet sili-
prevention of highway construction structures [205], the Cl cates such as clay minerals present as impurities) and alkali-
migration test, based on the method proposed by IBAC RWTH carbonate reaction (which involves the formation and expansion
Aachen [202] and the Chinese Standard GB/T 50082-2009 carbon- of brucite) represent less common forms of the alkali-aggregate
ation test [185]. reaction. The latter has been reported in dolomite-aggregate con-
taining shotcrete and concrete [17,212,213].
The use of the aggregates being quarried from tunnel excava-
7. Further durability issues tions for the concrete/shotcrete applied in the tunnels is becoming
a common practice [214–218]. Leemann et al. [219] reported a
This section covers four additional important aspects related to field case from a Swiss tunnel with 19–44 years old samples: even
shotcrete durability: alkali-aggregate reaction, shrinkage and though reactive aggregates were present and indication of alkali-
cracking, high temperature exposure, and microorganisms attack. silica reaction was detected, the damage was not considered sub-
The first one, alkali-aggregate reaction is especially relevant in stantial, and thus the same reactive aggregates were taken to be
excavation works, where the aggregates for the shotcrete are used for the shotcrete of new tunnels. These authors recommended
extracted on-site. The shotcrete resistance to most durability the use of pozzolanic SCMs and the installation of liners to prevent
issues is conditioned by the size, amount and distribution of cracks. contact with groundwater and subsequent alkali-aggregate reac-
Exposure to high temperature and fire is a key issue especially in tion. In relation with the use of accelerating admixtures, Won
tunnels. Finally, the interaction of cementitious materials with et al. [220] reported a correlation between expansion and alkali
microorganisms is a complex problem and not many studies are content in shotcrete with reactive aggregates. Aluminate-based
484 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

Fig. 9. Cracks formation in an underpass structure due to alkali-aggregate reaction, from [227] (a). Test panel with shotcrete protective layer after exposure to severe fire test
(b). Protective layer, which did not stand the same fire test as (b) followed by complete destruction of the layer (c). Shotcrete under microorganisms attack (d).

accelerators led to greater damage than alkali-free and calcium and, on a larger scale, the integrity between the shotcrete layers,
aluminate cement-based accelerators. Standards and methods cur- steel arches and bedrock gets destroyed.
rently used to assess the alkali reactivity of aggregates include e.g. The two main sources of shrinkage in concrete and shotcrete
petrographic examination of the aggregates (CSA A23.2-27A [280], are: (i) the smaller volume of the hydrated products, compared
Rilem method [221], and ASTM C295 [281]), accelerated mortar to the anhydrous components plus the water (so-called autoge-
bar testing (ASTM C1260 [282], ASTM C1567 [283], CSA A23.2- nous shrinkage), and (ii) the evaporation of water (so-called drying
25A [284] and NBRI test [222]), and concrete prism testing (ASTM shrinkage). Both mechanisms are more important in shotcrete than
C1293 [285] and CSA A23.2-14A [286]). An overview about the in concrete mainly because of the rapid stiffening and the higher
existing accelerated test methods for alkali-aggregate reactivity, cement content [227]. Furthermore, shrinkage in concrete is partly
the engineering implications of the alkali-aggregate reaction in compensated during setting. The main variables affecting the
concrete and how to prevent damage due to alkali-aggregates reac- extent of shrinkage in shotcrete are the mix constituents, ambient
tion is provided by Berube and Fournier, Fournier and Berube, and conditions [228], layer thickness [229], curing [230], and time
Thomas et al. [223–225]. Most recommendations to mitigate between successive spraying layers. Underlying material, from pre-
alkali-aggregate reaction risk include: limiting the alkali content viously sprayed layers or already in place material, restrains drying
of the concrete, and the use of fly ash, granulated blast furnace slag shrinkage, but it leads to heterogeneous curing and differential
and silica fume [226]. If reactive aggregate is used, the level of pre- shrinkage, which can then result in varying extent and intensity
vention has to be determined according to the reactivity of the of cracks formation. The type and dosage of accelerators also influ-
aggregate, the nature of the structure and the exposure conditions. ence the shrinkage behavior of shotcrete [164,231]. Alkali-free
accelerators typically result in larger shrinkage, as compared with
7.2. Shrinkage and cracking water glass-based setting accelerators [230]. Finally, higher
rebound leads to higher shrinkage, because the aggregates are
Shrinkage phenomena in shotcrete are relevant to durability, preferentially rebounding relative to the cement paste during the
because they may lead to the formation of cracks, which then make shotcreting process, causing an increase of the paste to aggregate
the shotcrete more vulnerable for external attacks such as leach- ratio in the fresh shotcrete, and thus a greater shrinkage.
ing, carbonation or SO2
4 ingress. In most cases, shrinkage cracking The use of reinforcement or the addition of fibers are proposed
is coupled with other tension sources, such as temperature gradi- as means of controlling shrinkage cracks in shotcrete. According to
ents and/or tensile loads. Other possible causes of cracking in shot- Leung et al. [232], shrinkage properties of wet-mixed shotcrete
crete include early deformations imposed by the rock mass. reinforced with either steel or PP fibers are very similar to those
Shrinkage and cracking are especially relevant in Scandinavian of concrete with a similar composition and reinforced with the
hard rock tunnels. The advanced stage of shrinkage-related crack- same type and amount of fibers. Other authors found no differ-
ing is expressed by the loss of mechanical performance; in partic- ences in shrinkage between plain and steel reinforced shotcrete
ular, the bond between the cement paste and aggregates gets lost [229]. Mixing steel and glass fibres has also been reported as a
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 485

way to prevent the development of shrinkage-related macrocracks towards other physical and chemical degradation phenomena
[233]. The use of shrinkage-reducing admixtures, based on poly- [248–250]. This bio-deposition process has actually been proposed
oxyalkylene alkyl ether, may help reducing drying shrinkage and as a method for repairing and restoring buildings [251,252]. Also,
preventing restrained induced cracking [234]. Another recent pro- Kalhori and Bagherpour [253] discussed its potential use for
posal to compensate shrinkage is the incorporation of expansive improving mechanical properties and healing cracks in shotcrete.
calcium sulfoaluminate cements into the shotcrete mix [235–237]. On the other side, certain microorganisms (like some algae,
fungi and cyanobacteria) can have deleterious effects, such as aes-
7.3. Temperature/Fire thetic changes, indoor air quality problems and bio-deterioration
of shotcrete and concrete [254] (Fig. 9d). An example of severe
Shotcrete is non-combustible and has good fire-resistance prop- bio-deterioration of concrete is the microbially induced sulfuric
erties, however, the term ‘fire-resistance’ is normally applied to acid corrosion in sewerage systems [255–257]. In such environ-
structural elements and not to materials [161]. Accordingly, the ments, biodeterioration of shotcrete and concrete very often pro-
time at which an element stops to perform adequately during gresses by the formation of organic (acetic, lactic, butyric,
exposure to fire determines its fire resistance. When exposed to carboxylic) and inorganic (sulfuric, carbonic) acids that can rapidly
heat, the cement paste first expands due to thermal expansion, degrade the cement paste and the aggregates by intense leaching
but then shrinks due to the progressive loss of water (Fig. 9b). Port- at pH <3. ‘Biogenic sulfuric acid’ (H2SO4), generated by the oxida-
landite decomposes at 400–500 °C and if re-wetted, the CaO rehy- tion of aqueous H2S by sulfur-oxidizing bacteria, is one of the most
drates causing expansion and disruption of the shotcrete. The corrosive substances in wastewater pipes and sewer systems. It
build-up of steam within confined spaces (i.e. pores) creates pres- dissolves the cementitious phases and often results in the forma-
sure gradients leading to shotcrete breakage (Fig. 9c); simultane- tion of secondary, expansive gypsum and possibly ettringite. A
ously, the steel reinforcement loses its mechanical detailed description of the various microbial deterioration mecha-
characteristics. The breaking-off of parts of the shotcrete layer is nisms in shotcrete and concrete is given in Wei et al. [258].
termed ‘spalling’, which is classified in three categories: surface, Despite the very few references, tunnels and subsea sprayed
corners and explosive [238]. However, due to the complexity of concrete structures seem to be the most prone to suffer from
the individual (tunnel) system and the huge differences between bio-deterioration. In Norway, the action of Mn- and Fe-oxidising
the seat of the fire, the exact mechanisms of spalling and explo- bacteria has been observed in subsea reinforced shotcrete tunnels.
sions are not fully understood yet [239]. The local acidification of the shotcrete surface resulted in paste
The degree of water saturation of the pores and their size distri- degradation as well as corrosion of steel fibres [191,259]. As men-
bution, the thickness of the shotcrete layer, mechanical loads and tioned before, microorganism attack is, in most cases, accompanied
type of aggregates are among the main factors affecting the fire by other types of physical and chemical deterioration, like sulfate
resistance of shotcrete. The behaviour of shotcrete towards high attack. The diversity of bacteria, the coupling of attacks and the
temperature is not necessarily similar to concrete, because of the few studies published on the topic, make this durability issue an
smaller aggregates, higher cement content and differences in open research field with still many knowledge gaps to be filled.
microstructure. The use of shotcrete for structural applications In this respect, Grengg et al. [260] recently argued that, at present,
has not only been limited by durability concerns, but also by high no cement-based material exists that is suitable to withstand the
temperature behaviour uncertainties [240]. Lower thermal diffu- very aggressive conditions attributed to microbial induced corro-
sivity and higher insulation properties are attributed to shotcrete sion of shotcrete and concrete over its entire life. Thus, there is still
with respect to concrete [241]. Also, the higher porosity of shot- a huge demand for the development of new materials and compos-
crete created from the entrained air helps releasing pressure in ites (e.g. antibacteriostatic agents, geopolymer concrete) to be
the pores when exposed to high temperature. According to applied in acid environments.
Bamonte et al. [240], alkali accelerator containing shotcrete is
more heat sensitive than alkali-free ones, the latter retaining a bet-
ter insulation capability at high temperatures. However, shotcrete 8. Final remarks and conclusions
made with alkali accelerator behaves not worse than ordinary
concrete. The durability of shotcrete is affected by the same variables as
In tunnel works, where fire can lead to very dangerous situa- that of concrete (mix composition, microstructure, permeability,
tions, steel fibres are often used to increase the structural ductility compaction, curing, environment, etc.), plus those only relevant
and toughness of the shotcrete layer. PP fibres are also added to the to shotcrete, i.e. composition and amount of accelerator, rock and
shotcrete/concrete layer to increase its resistance to spalling [242– water in contact with the shotcrete and application (shooting).
246]. These polymeric fibres (length <1 cm) create a microcracks These variables can be classified as controllable and uncontrollable,
network during melting, which helps releasing the pore pressure. meaning whether human input can influence them or not. The con-
In existing tunnels, protective layers of shotcrete reinforced with trollable ones can ideally be adapted to improve durability proper-
PP fibres have been applied to improve the fire resistance [247]. ties. For example, adequate curing, rather uncommon in
As opposed to the macro-fibres, used to reinforce the tunnel shot- underground shotcrete structures, guarantees proper hydration
crete, the micro-fibres used to increase the fire resistance con- and more homogeneous matrices. In the same way, proper shoot-
tribute to a much lower extent to the rebound material, and they ing of the shotcrete can significantly increase its final quality. How-
do not pose such an important pollution risk. ever, to what extent shooting can be controlled remains unclear, as
it depends on personal skills as well as on machinery and even for
7.4. Microorganisms attack high qualified personal complete homogeneous shooting is still
challenging. New tools are being developed to monitor the spray-
Microorganisms (e.g. bacteria and microscopic fungi) can have ing process parameters and the equipment function [261]: these
very different impacts on shotcrete durability. On the one side, cer- should allow in the future for a better control and understanding
tain microbes (like bacillus sphaericus, sporosarcina pasteurii and of the shotcreting practice.
bacillus subtilis) induce CaCO3 precipitation, preferentially in The mix composition (water, aggregates, binder, admixtures
cracks and pores, and on the shotcrete surface, which can both sig- and additions) has of course a direct impact on the microstructure
nificantly reduce the permeability and enhance the resistance and permeability. Maintaining the fulfilment of mechanical
486 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

criteria, the mixes can be optimised to reduce attack vulnerability. phases present in hydrated cement systems, together with their
As an illustration, Table 1 shows standard shotcrete components, role in hardened shotcrete. Average percentages occurring in fully
their range of variability and how they affect the durability proper- hydrated shotcrete (older than 90 days) are given. The table
ties of the resulting mix. The purpose of the table is to give an over- includes hydration products and phases formed as a result of an
view of the most important agents and impacts and, as such, it shall attack.
not be seen as a thorough description or information compilation. As seen, most attacks involve dissolution and precipitation pro-
From the non-controllable variables, mainly the type of rock cesses. Portlandite is, in the majority of cases, the first to dissolve,
and related characteristics (porosity, leachability of SO2 4 and when exposed to waters (alkali hydroxides being already in the
CO2–
3 ions, etc.) and the discharge and composition of percolating pore solution), acting as the provider of Ca2+ ions for the formation
waters, as well as ambient environmental conditions (temperature, of secondary phases, and maintaining the pH at values around 12.4.
relative humidity, pCO2), are relevant for shotcrete durability. C-S-H also contributes with Ca2+ ions through incongruent dissolu-
Table 2 shows under which scenario each durability issue may tion. Once portlandite is completely consumed, the pH starts to
become important. As mentioned before, mechanical loads, cracks drop and the Ca/Si ratio of the C-S-H continues decreasing. Al3+
and poor workmanship will increase the vulnerability of the lining and SO24 come from the dissolution of AFt and AFm phases, while
versus all the attacks. Table 3 includes a summary of the main CO2–
3 ions may come from the aggregates and from AFm phases.

Table 1
Relation between shotcrete components and effects on durability properties.

Shotcrete components Types and ranges Effect on durability performance

Cement Wet mixes:
Low C3A content: lower AFm formation
250–500 kg/m3
Sulfate content: controls ratio AFt/AFm
CEM I, SR, CEM II/A-S
Minimum SO3/Al2O3 needed for thaumasite formation
Dry mixes:
Ca/Si ratio relevant for thaumasite
Cem I or ‘spray cement’ (low sulfate content)
High OPC content higher leaching and sintering potential (lower carbonation front
360–450 kg/m3 penetration)
Aggregates Calcite, Dolomite, Siliceous
Alkali-carbonate reaction
1500–1900 kg/m3
Alkali-silica reaction
Size 0–4 mm, 4–8 mm
Leaching of carbonates

Contribution to thaumasite formation

Freeze-thaw resistance
Accelerators 3–12 wt% of the binder
Changes in porosity and permeability
-Alkali-containing NaAlO2, Na2SiO3, Na2CO3
Control microstructure (e.g. SO3/Al2O3 ratio: AFt-AFm formation)
-Alkali-free Al2(SO4)3,
Al2(SO4)3 and NaAlO2 favour thaumasite formation
Ca(NO3)2
Reduce resistance to freeze thaw (carbonate-based especially)
-Cement-based calcium aluminate, calcium
Aluminate-based greater expansion in alkali-silica reaction than alkali-free
sulfoaluminate
Organics: effects on the durability of shotcrete are still unclear
Superplasticizers 0.5–1.5 wt% of the binder
Reduce water demand and porosity

Other direct effects on durability are unclear
Retarders 0–0.5 wt% of the binder
No obvious influence on durability properties
Water 180–210 kg/m3
Control porosity and degree of hydration
w/b 0.45–0.50
Higher porosity: higher risk for sulfate attack, leaching, etc.

More water, higher shrinkage
SCMs Granulated blast furnace slag, fly ash, silica fume,
Reduce porosity and Ca(OH)2
metakaolin
Reduce alkali-silica reaction, improve sulfate resistance
0–200 kg/m3
Improve leaching resistance and prevents sinter formation

Reduce carbonation front penetration resistance
Air-entraining agent 0.2–2% wt. % of the binder
Increases air voids: act as expansion reservoirs

Improve frost and deicing salt scaling resistance and provides space for newly
formed sulfate phases
Fibres Steel 20–60 kg/m3
Macro-fibres increase ductility (post crack behaviour)
PP-macro 3–5 kg/m3
Steel: corrosion risk
PP-micro 1–2.5 kg/m3
Micro-PP fibres: increase resistance to fire
Micro <0.3 mm diameter
Micro-PP: increase freeze-thaw resistance, reduce early shrinkage cracks
Macro >0.3 mm diameter

Table 2
Coupling of attacks with relevant conditions.

Attack Rock type Water Temperature Relative humidity Other factors

Leaching/soft water Silicic rocks e.g. Percolating All At low T calcite All usually high Pressure and fast flowing (e.g. head race
attack granite gneiss water solubility higher tunnels in power plants)
Sintering All types Drainage water All All usually high Degassing and CO2 absorption
Mixing of solutions can be critical
Sulfate attack with Marine evaporates Groundwater Low (5–15 °C) All Cyclic and evaporation High SO2
4 concentration
thaumasite formation (gypsum) especially dangerous
Sulfide oxidation
Freeze-thaw High saturation Low and cycles High Chloride ions
High temperature
gradients
Steel corrosion Marine evaporates Chloride and CO2 All Intermediate Deicing salts, CO2 from cars exhaust gas
(salt) containing
Alkali-aggregate Marine evaporates High saturation All High Alkalis from seawater, deicing salts
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 487

Table 3
Relevant phases in shotcrete fully hydrated matrices with alkali-free Al2(SO4)3-based accelerator.

Phase Content in the hydrated matrix Role/relevance in hardened shotcrete Consequences

Portlandite 15–20 wt% of the paste (with SCMs High pH maintenance: protection against steel corrosion CaCO3 formation: block of
reduced down to 5) Leaching drainage systems
Source of Ca2+ for carbonation, sintering, etc.
C-S-H 40–70 wt% of the paste Leaching Strength loss
Sintering
Carbonation
Source of silica for thaumasite formation
Ettringite 5–15 wt% of the paste Hydration product: conversion to AFm Secondary formation of ettringite:
Damage product from sulfate attack expansion
AFm phases 5–10 wt% of the paste Cl- AFm: Chloride combination Delays electrochemical corrosion
SO4-AFm: Conversion to ettringite Formation of expansive ettringite
Calcite Up to 60% of shotcrete (aggregates) Inert/reactive component of shotcrete matrix May increase risk of thaumasite
(dolomite) Calcite: up to 2 wt% of the paste formation
(cement)
Calcite: damage product from carbonation and sintering Calcite: main sinter product, block
of drainage systems
Gypsum Damage product from sulfate attack Cracking, expansion
Thaumasite Damage product from sulfate attack Loss of strength, softening
Brucite Damage product from magnesium sulfate attack Damage product from Expansion, cracking
alkali-carbonate (dolomite) reaction
Mirabilite, Damage product from efflorescence type of sulfate attack Expansion and physical damage
thenardite

The other key parameters, porosity and permeability, are worse in horizontal parts, where water will accumulate, thus
affected by most of the variables involved in the process of making reaching higher water saturation levels.
concrete or shotcrete. A direct comparison between shotcrete and The inhomogeneity of shotcrete matrices should also be consid-
concrete is difficult because of the broad ranges of composition ered when extracting cores and designing durability experiments.
variability and the differences in uses. The ‘zero-concrete’, a refer- Average values from various zones may not be helpful to evaluate
ence concrete with the same composition as the shotcrete with the the performance of poor layers.
exception of the accelerator, absent in the reference concrete, is the More research is needed to study the evolution of the
closest concrete-shotcrete equivalent composition. This reference microstructure of shotcrete matrices during hydration from early
concrete is used to evaluate changes in technological properties, ages to fully hydrated stages, in parallel to its evolution under
e.g. loss of strength, directly related to the porosity. However, in aggressive conditions in order to be able to deconvolute both,
this case the differences in the pores development (and and understand the phases’ roles and their interactions in the var-
microstructure) are the result of the combined effect of the accel- ious reactions. At the same time, generic experiments should be
erator and the shooting. designed to better understand degradation processes in terms of
The tests and methods used in the lab to measure resistance of dissolution and precipitation rates.
shotcrete to aggressive environments lack the same ‘reality’ as the In-situ monitoring appears as an essential tool to follow dam-
ones used for concrete with the extra difficulty to prepare testing age of shotcrete and understand mechanisms and kinetics. The
specimens. This latter involves two aspects: (i) reproducing spray- coupling of experimental data with equilibrium calculations is also
ing conditions in a lab scale and (ii) spraying boxes on-site. In the needed to evaluate saturation indexes and to be able to model and
first case, special lab-scale devices are needed, like the MiniShot, predict potential risks.
developed by Sika [262], or the one described in [263], but only a Finally, another interesting path to follow is to look at examples
few of these exist and availability is restricted. The second option, of existing durable shotcrete structures, as ‘the best test of durabil-
spraying testing panels directly after real-scale spraying, poses ity is the test of time’ [264]. The characteristics of the durable shot-
advantages in terms of ‘reality’, but the shooting of homogeneous crete structures, together with the exposure conditions they have
layers becomes even more challenging over small surfaces than in been subjected to, should be obtained to extract the relevant infor-
vast areas. This means, differences (in homogeneity) between real mation on how to design appropriate mixes for specific sites. Nev-
shotcrete and cores extracted from sprayed boxes are larger than ertheless, the shotcrete from the old durable structures may have
the ones between standard prisms and real non-sprayed concrete. few in common with modern shotcrete mixes used nowadays in
A few methods have been developed recently to assess durabil- tunnelling works, and the lessons learned may not be directly
ity of shotcrete, like the RV test, but if the objective is to achieve transferable.
long-term durable shotcrete structures more reliable methods The future shotcrete should last as long as concrete, but it
have to be proposed. In parallel, the criteria for acceptable perfor- should also contribute to the reduction of raw materials demand
mance should be correlated with service life predictions. Layer and CO2 emissions. To reach that goal, both scientific and technical
thickness, usually between 2 and 10 cm in northern countries inputs are needed. Newly developed mixes with low clinker con-
(hard rock) and around 30 cm in the Alps region (soft rock and tent, suitable SCMs and optimised mechanical properties need to
soil), may be then optimized for price and performance. be assessed. The role of fillers and admixtures on the durability
Specific conditions should be reproduced in lab tests to achieve of these new mixes is still not fully clear, and it will play a major
more realistic results. Geometric considerations should also be role in the design of shotcrete in the next years.
included for specific testing, for example, where different flow
regimes are established in horizontal and vertical elements. Dura-
bility issues like leaching will be more intensive in vertical ele- Conflict of interest
ments, especially in the upper parts, as water will flow down to
the lower parts; frost damage risk, on the other hand, may be The authors declare no conflict of interest.
488 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

Acknowledgements [28] D. Planel, J. Sercombe, P. Le Bescop, F. Adenot, J.-M. Torrenti, Long-term

performance of cement paste during combined calcium leaching–sulfate
attack: kinetics and size effect, Cem. Concr. Res. 36 (2006) 137–143, https://
The authors gratefully acknowledge the funding by the industry [Link]/10.1016/[Link].2004.07.039.
partners, the Austrian Research Promotion Agency FFG (ASSpC [29] B. Bary, Simplified coupled chemo-mechanical modeling of cement pastes
behavior subjected to combined leaching and external sulfate attack, Int. J.
Project-No. 856080) and the Austrian Building Technology Associ-
Numer. Anal. Methods Geomech. 32 (2008) 1791–1816, [Link]
ation ÖBV. The colleagues from the ASSpC project are thanked for 10.1002/nag.696.
the fruitful discussions. Günther Koraimann, Christian Neumann, [30] C. Rogers, R. Hooton, C. Rogers, R. Hooton, Reduction in Mortar and Concrete
Expansion with Reactive Aggregates Due to Alkali Leaching, in: Cem. Concr.
Eric Bauer and Benedikt Lindlar are also thankfully acknowledged
Aggregates. 13 (1991) 42–49, [Link]
for granting permission to use their images in the paper. [31] S. Multon, A. Sellier, Multi-scale analysis of alkali-silica reaction (ASR):
Impact of alkali leaching on scale effects affecting expansion tests, Cem.
Concr. Res. 81 (2016), [Link]
References [32] P. Rivard, M.A. Bérubé, J.P. Ollivier, G. Ballivy, Decrease of pore solution
alkalinity in concrete tested for alkali-silica reaction, Mater. Struct. Constr. 40
(2007), [Link]
[1] EFNARC, European Specification for Sprayed Concrete, Surrey, 1996.
[33] S. Chatterji, The role of Ca(OH)2 in the breakdown of Portland cement
[2] EFNARC, European Specification for sprayed concrete. Guidelines for
concrete due to alkali-silica reaction, Cem. Concr. Res. 9 (1979), [Link]
specifiers and contractors, Surrey, 1999.
org/10.1016/0008-8846(79)90024-3.
[3] M. Jolin, D. Beaupré, Understanding wet-mix shotcrete: mix design,
[34] W. Müllauer, R.E. Beddoe, D. Heinz, Effect of carbonation, chloride and
specifications, and placement, Shotcrete Mag. Summer 2003, Am. Shotcrete
external sulphates on the leaching behaviour of major and trace elements
Assoc. (2003).
from concrete, Cem. Concr. Compos. 34 (2012), [Link]
[4] W. Kusterle, C. Neumann, J. Max, J. Jäger and R. Röck, Spritzbeton im
cemconcomp.2012.02.002.
Tunnelbau, in: K. Bergmeister, F. Fingerloos and J.-.D. Wömer, (Eds.),
[35] V.B. Duong, R. Sahamitmongkol, S. Tangtermsirikul, Effect of leaching on
BetonKalender, Unterirdisches Bauen, Grundbau, Eurocode 7, 2014, Ernst &
carbonation resistance and steel corrosion of cement-based materials, Constr.
Sohn; Berlin.
Build. Mater. 40 (2013), [Link]
[5] Japan International Cooperation Agency and Road Development Authority,
[36] Y. Qin, H. Yang, Carbonation dominates the acid intake of recycled concrete
Guideline for Design of Road Tunnel, 2018.
aggregate subjected to intermittent leaching, Constr. Build. Mater. 89 (2015),
[6] Norwegian Concrete Association, Sprayed Concrete for Rock Support, 2011.
[Link]
[7] Australian Shotcrete Society, Shotcreting in Australia - recommended
[37] L. Mayer, Beton - keine Gefahr für Boden und Grundwasser, Beton-und
practice, 2010.
Stahlbetonbau. Heft 3 (1994).
[8] American Concrete Institute Commitee 506, Guide to Shotcrete, 2016.
[38] J. Bensted, A.R. Brough, M.M. Page, Chemical degradation of concrete, in: C.L.
[9] Austrian Building Technology Association (ÖBV), Guideline: Sprayed
Page, M.M. Page (Eds.), Durab. Concr. Cem. Compos., Woodhead Publishing,
Concrete, Vienna, 2013.
Birmingham, 2007.
[10] EN 14487–1, Sprayed concrete. Definitions, specifications and conformity,
[39] J. Duchesne, A. Bertron, Leaching of cementitious materials by pure water and
Brussels, 2005.
strong acids (HCl and HNO3), in: Perform. Cem. Mater. Aggress. Aqueous
[11] EN 206–1, Concrete – Part 1: Specification, performance, production and
Environ. State-of-the-Art Report, RILEM TC 211 - PAE, 2013.
conformity, Brussels, 2014.
[40] J. Cowie, F.P. Glasser, The reaction between cement and natural waters
[12] A. Thomas, Sprayed Concrete Lined Tunnels, CRC Taylor and Francis Group,
containing dissolved carbon dioxide, Adv. Cem. Res. 4 (1992) 119–134,
2008.
[Link]
[13] A. Thomas and R. Dimmock, The design philosophy for permanent sprayed
[41] J. Jain, N. Neithalath, Analysis of calcium leaching behavior of plain and
concrete linings, Eighth Int. Symp. Sprayed Concr., 2018; Trondheim, Norway.
modified cement pastes in pure water, Cem. Concr. Compos. 31 (2009),
[14] L. Rabcewicz, J. Gosler, Application of the NATM to the underground works at
[Link]
Tarbela. Water Power 26, 1974.
[42] B. Lothenbach, K. Scrivener, R.D. Hooton, Supplementary cementitious
[15] T. Cordes, M. Hofmann, R. Murr and K. Bergmeister, Current developments in
materials, Cem. Concr. Res. 41 (2011), [Link]
shotcrete technology at the Brenner Base Tunnel, in: W. Kusterle, (Ed),
cemconres.2010.12.001.
Spritzbet. Tagung, 2018; Alpbach, Austria.
[43] R.J. Van Eijk, H.J.H. Brouwers, Study of the relation between hydrated
[16] F. Mittermayr, A. Baldermann, C. Kurta, T. Rinder, D. Klammer, A. Leis, J.
portland cement composition and leaching resistance, Cem. Concr. Res. 28
Tritthart, M. Dietzel, Evaporation - a key mechanism for the thaumasite form
(1998).
of sulfate attack, Cem. Concr. Res. 49 (2013), [Link]
[44] M. Mainguy, O. Coussy, Propagation fronts during calcium leaching and
cemconres.2013.03.003.
chloride penetration, J. Eng. Mech. 126 (2000), [Link]
[17] F. Mittermayr, A. Baldermann, C. Baldermann, G.H. Grathoff, D. Klammer, S.J.
0733-9399(2000) 126:3(250).
Köhler, A. Leis, L.N. Warr, M. Dietzel, Environmental controls and reaction
[45] B. Johannesson, K. Yamada, L.-O. Nilsson, Y. Hosokawa, Multi-species ionic
pathways of coupled de-dolomitization and thaumasite formation, Cem.
diffusion in concrete with account to interaction between ions in the pore
Concr. Res. 95 (2017), [Link]
solution and the cement hydrates, Mater. Struct. Constr. 40 (2007), https://
[18] L.R. Prudêncio Jr., Accelerating admixtures for shotcrete, Cem. Concr. Compos.
[Link]/10.1617/s11527-006-9176-y.
20 (1998) 213–219.
[46] F.P. Glasser, J. Marchand, E. Samson, Durability of concrete - Degradation
[19] [Link], R.A. Kaden, Durability of Shotcrete, Spec. Publ. 100, American
phenomena involving detrimental chemical reactions, Cem. Concr. Res. 38
Concrete Institute, 1987, 1071-1101. doi:10.14359/3744.
(2008), [Link]
[20] M. Lundgren, E. Helsing, A. Babaahmadi, U. Mueller, State-of-the-Art Report
[47] P. Feng, C. Miao, J.W. Bullard, A model of phase stability, microstructure and
on: Material Type, Requirements and Durability aspects of Sprayed Concrete
properties during leaching of portland cement binders, Cem. Concr. Compos.
in Tunnels, CBI Swedish Cement and Concrete Research Institute, Built
49 (2014), [Link]
Environment, RISE - Research Institutes of Sweden (2018).
[48] D. Jacques, L. Wang, E. Martens, D. Mallants, Modelling chemical degradation
[21] P. Blanc, X. Bourbon, A. Lassin, E.C. Gaucher, Chemical model for cement-
of concrete during leaching with rain and soil water types, Cem. Concr. Res.
based materials: thermodynamic data assessment for phases other than C-S-
40 (2010), [Link]
H, Cem. Concr. Res. 40 (2010) 1360–1374, [Link]
[49] E. Stora, B. Bary, Q.-C. He, E. Deville, P. Montarnal, Modelling and simulations
cemconres.2010.04.003.
of the chemo-mechanical behaviour of leached cement-based materials:
[22] N.C.M. Marty, O. Bildstein, P. Blanc, F. Claret, B. Cochepin, E.C. Gaucher, D.
Interactions between damage and leaching, Cem. Concr. Res. 40 (2010),
Jacques, J.-E. Lartigue, S. Liu, K.U. Mayer, J.C.L. Meeussen, I. Munier, I.
[Link]
Pointeau, D. Su, C.I. Steefel, Benchmarks for multicomponent reactive
[50] C.J. Engelsen, H.A. Van Der Sloot, G. Wibetoe, H. Justnes, W. Lund, E.
transport across a cement/clay interface, Comput. Geosci. 19 (2015) 635–
Stoltenberg-Hansson, Leaching characterisation and geochemical modelling
653, [Link]
of minor and trace elements released from recycled concrete aggregates,
[23] H.F.W. Taylor, Cement Chemistry, Thomas Telford, 1997.
Cem. Concr. Res. 40 (2010), [Link]
[24] P. Faucon, F. Adenot, M. Jorda, R. Cabrillac, Behaviour of crystallised phases of
[51] B. Gérard, C. Le Bellego, O. Bernard, Simplified modelling of calcium leaching
Portland cement upon water attack, Mater. Struct. 30 (1997) 480–485,
of concrete in various environments, Mater. Struct. Constr. 35 (2002), https://
[Link]
[Link]/10.1617/13937.
[25] M. Rosenqvist, A. Bertron, K. Fridh, M. Hassanzadeh, Concrete alteration due
[52] A. Hidalgo, S. Petit, C. Domingo, C. Alonso, C. Andrade, Microstructural
to 55 years of exposure to river water: chemical and mineralogical
characterization of leaching effects in cement pastes due to neutralisation of
characterisation, Cem. Concr. Res. 92 (2017) 110–120, [Link]
their alkaline nature. Part I: portland cement pastes, Cem. Concr. Res. 37
10.1016/[Link].2016.11.012.
(2007), [Link]
[26] R.S. Gollop, H.F.W. Taylor, Microstructural and microanalytical studies of
[53] K. Nakarai, T. Ishida, K. Maekawa, Modeling of calcium leaching from cement
sulfate attack. I. Ordinary portland cement paste, Cem. Concr. Res. 22 (1992),
hydrates coupled with micro-pore formation, J. Adv. Concr. Technol. 4 (2006),
[Link]
[Link]
[27] E. Rozière, A. Loukili, R. El Hachem, F. Grondin, Durability of concrete exposed
[54] T. De Larrard, F. Benboudjema, J.B. Colliat, J.M. Torrenti, F. Deleruyelle,
to leaching and external sulphate attacks, Cem. Concr. Res. 39 (2009) 1188–
Concrete calcium leaching at variable temperature: experimental data and
1198, [Link]
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 489

numerical model inverse identification, Comput. Mater. Sci. 49 (2010), [81] J.J. Chen, J.J. Thomas, H.M. Jennings, Decalcification shrinkage of cement
[Link] paste, Cem. Concr. Res. 36 (2006), [Link]
[55] D. Gawin, F. Pesavento, B.A. Schrefler, Modeling of cementitious materials cemconres.2005.11.003.
exposed to isothermal calcium leaching, considering process kinetics and [82] K. Yokozeki, K. Watanabe, N. Sakata, N. Otsuki, Modeling of leaching from
advective water flow. Part 2: numerical solution, Int. J. Solids Struct. 45 cementitious materials used in underground environment, Appl. Clay Sci. 26
(2008), [Link] (2004), [Link]
[56] D. Gawin, F. Pesavento, B.A. Schrefler, Modeling of cementitious materials [83] C.J. Engelsen, H.A. van der Sloot, G. Wibetoe, G. Petkovic, E. Stoltenberg-
exposed to isothermal calcium leaching, considering process kinetics and Hansson, W. Lund, Release of major elements from recycled concrete
advective water flow. Part 1: theoretical model, Int. J. Solids Struct. 45 (2008), aggregates and geochemical modelling, Cem. Concr. Res. 39 (2009), https://
[Link] [Link]/10.1016/[Link].2009.02.001.
[57] D. Gawin, F. Pesavento, B.A. Schrefler, Modeling deterioration of cementitious [84] H.A. Van Der Sloot, Comparison of the characteristic leaching behavior of
materials exposed to calcium leaching in non-isothermal conditions, Comput. cements using standard (EN 196–1) cement mortar and an assessment of
Methods Appl. Mech. Eng. 198 (2009), [Link] their long-term environmental behavior in construction products during
cma.2009.05.005. service life and recycling, Cem. Concr. Res. 30 (2000), [Link]
[58] F.-J. Ulm, E. Lemarchand, F.H. Heukamp, Elements of chemomechanics of S0008-8846(00)00287-8.
calcium leaching of cement-based materials at different scales, Eng. Fract. [85] NEN 7341:1995 Leaching Characteristics Of Solid Earthy And Stony Building
Mech. 70 (2003), [Link] And Waste Materials - Leaching Tests - Determination Of The Availability Of
[59] C. Le Bellégo, G. Pijaudier-Cabot, B. Gérard, J.-F. Dubé, L. Molez, Coupled Inorganic Components For Leaching, Delft, 1995.
mechanical and chemical damage in calcium leached cementitious [86] W. Kusterle, W. Pichler, A. Saxer, Reduced leaching behaviour of sprayed
structures, J. Eng. Mech. 129 (2003), [Link] concrete (in German), Beton-und Stahlbetonbau. 106 (2011) 847–852,
9399(2003) 129:3(333). [Link]
[60] F. Bernard, S. Kamali-Bernard, W. Prince, 3D multi-scale modelling of [87] M. Thumann, M. Astner, A. Saxer, W. Kusterle, Precipitations in the Tunnel
mechanical behaviour of sound and leached mortar, Cem. Concr. Res. 38 Drainage System – Optimized Shotcrete Mix-Design, in: Shotcrete Undergr.
(2008), [Link] Support XII, Singapore, 2015.
[61] D. Kuhl, F. Bangert, G. Meschke, Coupled chemo-mechanical deterioration of [88] B. Aitchison, D. Oliveira, D. Backhouse and R. Netterfield, Design and
cementitious materials. Part I: modeling, Int. J. Solids Struct. 41 (2004), construction of a shallow cover tunnel junction using sprayed concrete,
[Link] Sydney Australia, in: T. Beck, S. Myren and S. Engen, (Eds.), Eighth Int. Symp.
[62] F.H. Heukamp, F.-J. Ulm, J.T. Germaine, Mechanical properties of calcium- Sprayed Concr., 2018; Trondheim, Norway.
leached cement pastes: triaxial stress states and the influence of the pore [89] A. Ansell and L.E. Bryne, In-situ observations and laboratory testing of
pressures, Cem. Concr. Res. 31 (2001), [Link] shrinkage cracking in shotcrete on soft drains, in: W. Kusterle, (Ed), Spritzbet.
(01)00472-0. Tagung, 2015; Alpbach, Austria.
[63] V.H. Nguyen, H. Colina, J.M. Torrenti, C. Boulay, B. Nedjar, Chemo-mechanical [90] M.R.M. Saade, A. Passer, F. Mittermayr, A Preliminary Systematic
coupling behaviour of leached concrete. Part I: experimental results, Nucl. Investigation onto Sprayed Concrete’s Environmental Performance, Procedia
Eng. Des. 237 (2007), [Link] CIRP. 69 (2018) 212–217, [Link]
[64] V.H. Nguyen, B. Nedjar, J.M. Torrenti, Chemo-mechanical coupling behaviour [91] T. Gamisch and G., Girmscheid, Versinterungsprobleme in
of leached concrete. Part II: Modelling, Nucl. Eng. Des. 237 (2007), [Link] Bauwerksentwässerungen (in German), Bauwerk, Verlag, 2007.
org/10.1016/[Link].2007.02.012. [92] Austrian Concrete and Building Technology Association (ÖVBB), Guideline:
[65] F.-J. Ulm, J.-M. Torrenti, F. Adenot, Chemoporoplasticity of calcium leaching in Tunnel Drainage (in German), Vienna, 2010.
concrete, J. Eng. Mech. 125 (1999). [93] M. Dietzel, E. Usdowski, J. Hoefs, Chemical and 13C/12C- and 18O/16O-
[66] J.M. Torrenti, V.H. Nguyen, H. Colina, F. Le Maou, F. Benboudjema, F. isotope evolution of alkaline drainage waters and the precipitation of calcite,
Deleruyelle, Coupling between leaching and creep of concrete, Cem. Concr. Appl. Geochem. 7 (1992), [Link]
Res. 38 (2008), [Link] [94] M. Dietzel, B. Purgstaller, A. Leis, P. Reichl, H. Stadler, A. Niedermayr, T.
[67] C. Le Bellégo, B. Gérard, G. Pijaudier-Cabot, Chemo-mechanical effects in Rinder, H. Wagner, Current challenges for scaling of tunnel drainage systems
mortar beams subjected to water hydrolysis, J. Eng. Mech. 126 (2000), - Modelling approaches, monitoring tools and prevention strategies,
[Link] 126:3(266). Geomech. Und Tunnelbau. 6 (2013), [Link]
[68] C. Carde, R. François, Aging damage model of concrete behavior during the [95] R. Boch, M. Dietzel, P. Reichl, A. Leis, A. Baldermann, F. Mittermayr, P. Pölt,
leaching process, Mater. Struct. Constr. 30 (1997). Rapid ikaite (CaCO36H2O) crystallization in a man-made river bed:
[69] K. Wang, D.C. Jansen, S.P. Shah, A.F. Karr, Permeability study of cracked Hydrogeochemical monitoring of a rarely documented mineral formation,
concrete, Cem. Concr. Res. 27 (1997). Appl. Geochem. 63 (2015), [Link]
[70] P. Locoge, M. Massat, J.P. Ollivier, C. Richet, Ion diffusion in microcracked [96] L.P. Field, A.E. Milodowski, R.P. Shaw, L.A. Stevens, M.R. Hall, A. Kilpatrick, J.
concrete, Cem. Concr. Res. 22 (1992), [Link] Gunn, S.J. Kemp, M.A. Ellis, Unusual morphologies and the occurrence of
90085-A. pseudomorphs after ikaite (CaCO3/6H2O) in fast growing, hyperalkaline
[71] B. Gérard, J. Marchand, Influence of cracking on the diffusion properties of speleothems, Mineral. Mag. 81 (2017) 565–589, [Link]
cement-based materials. Part I: influence of continuous cracks on the steady- minmag.2016.080.111.
state regime, Cem. Concr. Res. 30 (2000), [Link] [97] A. Niedermayr, S.J. Köhler, M. Dietzel, Impacts of aqueous carbonate
(99)00201-X. accumulation rate, Magnesium and polyaspartic acid on calcium carbonate
[72] C. Vernet, C. Alonso, C. Andrade, M. Castellote, I. Llorente, A. Hidalgo, A new formation (6–40°C), Chem. Geol. 340 (2013), [Link]
leaching test based in a running water system to evaluate long-term water chemgeo.2012.12.014.
resistance of concretes, Adv. Cem. Res. 14 (2002), [Link] [98] B. Purgstaller, F. Konrad, M. Dietzel, A. Immenhauser, V. Mavromatis, Control
adcr.14.4.157.38912. of Mg2+/Ca2+Activity Ratio on the Formation of Crystalline Carbonate
[73] C. Alonso, M. Castellote, I. Llorente, C. Andrade, Ground water leaching Minerals via an Amorphous Precursor, Cryst. Growth Des. 17 (2017) 1069–
resistance of high and ultra high performance concretes in relation to the 1078, [Link]
testing convection regime, Cem. Concr. Res. 36 (2006), [Link] [99] M. Dietzel, T. Rinder, A. Niedermayr, F. Mittermayr, L. A., D. Klammer, S.
10.1016/[Link].2006.04.004. Köhler, P. Reichl,, Ursachen und Mechanismen der Versinterung von
[74] E. Rozière, A. Loukili, Performance-based assessment of concrete resistance to Tunneldrainagen, BHM Berg- Und Hüttenmännische, Monatshefte. 153
leaching, Cem. Concr. Compos. 33 (2011), [Link] (2008) 369–372, [Link]
cemconcomp.2011.02.002. [100] T. Rinder, M. Dietzel, A. Leis, Calcium carbonate scaling under alkaline
[75] EN 14429:2015, Characterisation of waste – Leaching behaviour tests – conditions - case studies and hydrochemical modelling, Appl. Geochem. 35
Influence of pH on leaching with initial acid/base addition, Brussels, 2015. (2013), [Link]
[76] M. Castellote, I. Llorente, C. Andrade, C. Alonso, Accelerated leaching of ultra [101] Austrian Building Technology Association (ÖBV), Guideline: Sintering
high performance concretes by application of electrical fields to simulate potential determination (in German), Vienna, 2012.
their natural degradation, Mater. Struct. Constr. 36 (2003), [Link] [102] M. Whittaker, L. Black, Current knowledge of external sulfate attack, Adv.
10.1617/13795. Cem. Res. 27 (2015), [Link]
[77] H. Saito, S. Nakane, Comparison between diffusion test and electrochemical [103] W. Kunther, B. Lothenbach, K.L. Scrivener, On the relevance of volume
acceleration test for leaching degradation of cement hydration products, ACI increase for the length changes of mortar bars in sulfate solutions, Cem.
Mater. J. 96 (1999). Concr. Res. 46 (2013), [Link]
[78] H. Saito, A. Deguchi, Leaching tests on different mortars using accelerated [104] W. Müllauer, R.E. Beddoe, D. Heinz, Sulfate attack expansion mechanisms,
electrochemical method, Cem. Concr. Res. 30 (2000), [Link] Cem. Concr. Res. 52 (2013) 208–215, [Link]
S0008-8846(00)00377-X. cemconres.2013.07.005.
[79] C. Carde, G. Escadeillas, R. François, Use of ammonium nitrate solution to [105] T. Schmidt, B. Lothenbach, M. Romer, J. Neuenschwander, K. Scrivener,
simulate and accelerate the leaching of cement pastes due to deionized Physical and microstructural aspects of sulfate attack on ordinary and
water, Mag. Concr. Res. 49 (1997). limestone blended Portland cements, Cem. Concr. Res. 39 (2009), [Link]
[80] C. Gallé, H. Peycelon, P. Le Bescop, Effect of an accelerated chemical org/10.1016/[Link].2009.08.005.
degradation on water permeability and pore structure of cement-based [106] S.M. Torres, C.J. Lynsdale, R.N. Swamy, J.H. Sharp, Microstructure of 5-year-
materials, Adv. Cem. Res. 16 (2004), [Link] old mortars containing limestone filler damaged by thaumasite, Cem. Concr.
adcr.16.31.105.41515. Res. 36 (2006), [Link]
490 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

[107] S.T. Lee, R.D. Hooton, H.-S. Jung, D.-H. Park, C.S. Choi, Effect of limestone filler sulfate attack and drying-wetting cycles, Constr. Build. Mater. 81 (2015),
on the deterioration of mortars and pastes exposed to sulfate solutions at [Link]
ambient temperature, Cem. Concr. Res. 38 (2008), [Link] [135] J. Wang, D. Niu, Influence of freeze–thaw cycles and sulfate corrosion
cemconres.2007.08.003. resistance on shotcrete with and without steel fiber, Constr. Build. Mater. 122
[108] H. Justnes, Thaumasite formed by sulfate attack on mortar with limestone (2016), [Link]
filler, Cem. Concr. Compos. 25 (2003), [Link] [136] J. Wang, D. Niu, Z. Song, Damage layer thickness and formation mechanism of
(03)00120-3. shotcrete with and without steel fiber under sulfate corrosion of dry–wet
[109] E.F. Irassar, V.L. Bonavetti, M. González, Microstructural study of sulfate cycles by ultrasound plane testing method, Constr. Build. Mater. 123 (2016),
attack on ordinary and limestone Portland cements at ambient temperature, [Link]
Cem. Concr. Res. 33 (2003), [Link] [137] S. Lorente, M.-P. Yssorche-Cubaynes, J. Auger, Sulfate transfer through
[110] S. Tsivilis, K. Sotiriadis, A. Skaropoulou, Thaumasite form of sulfate attack concrete: migration and diffusion results, Cem. Concr. Compos. 33 (2011),
(TSA) in limestone cement pastes, J. Eur. Ceram. Soc. 27 (2007), [Link] [Link]
org/10.1016/[Link].2006.05.048. [138] K. van Tittelboom, N. de Belie, D. Hooton, Test Methods for Resistance of
[111] A.M. Ramezanianpour, R.D. Hooton, Thaumasite sulfate attack in Portland Concrete to Sulfate Attack - A Critical Review, in: M. Alexander, A. Bertron, N.
and Portland-limestone cement mortars exposed to sulfate solution, Constr. De Belie (Eds.), Perform. Cem. Mater. Aggress. Aqueous Environ. - State-of-
Build. Mater. 40 (2013), [Link] the-Art Report, RILEM TC 211 - PAE, Springer Netherlands, 2013.
[112] A. Skaropoulou, K. Sotiriadis, G. Kakali, S. Tsivilis, Use of mineral admixtures [139] E. Freyburg, A.M. Berninger, Field experiences in concrete deterioration by
to improve the resistance of limestone cement concrete against thaumasite thaumasite formation: possibilities and problems in thaumasite analysis,
form of sulfate attack, Cem. Concr. Compos. 37 (2013), [Link] Cem. Concr. Compos. 25 (2003), [Link]
10.1016/[Link].2013.01.007. 00135-5.
[113] E.F. Irassar, Sulfate attack on cementitious materials containing limestone [140] P. Hagelia, R.G. Sibbick, N.J. Crammond, C.K. Larsen, Thaumasite and
filler - a review, Cem. Concr. Res. 39 (2009), [Link] secondary calcite in some Norwegian concretes, Cem. Concr. Compos. 25
cemconres.2008.11.007. (2003), [Link]
[114] A. Baldermann, M. Rezvani, T. Proske, C. Grengg, F. Steindl, M. Sakoparnig, C. [141] P. Hagelia, Deterioration Mechanisms and Durability of Sprayed Concrete for
Baldermann, I. Galan, F. Emmerich, F. Mittermayr, Effect of very high Rock Support in Tunnels, University of Oslo, Norway, 2011.
limestone content and quality on the sulfate resistance of blended [142] G.-C. Long, Y.-J. Xie, D.-H. Deng, X.-K. Li, Deterioration of concrete in railway
cements, Constr. Build. Mater. 188 (2018), [Link] tunnel suffering from sulfate attack, J. Cent. South Univ. Technol. 18 (2011),
[Link].2018.08.169. [Link]
[115] N.J. Crammond, The thaumasite form of sulfate attack in the UK, Cem. Concr. [143] M. Romer, Steam locomotive soot and the formation of thaumasite in
Compos. 25 (2003), [Link] shotcrete, Cem. Concr. Compos. 25 (2003), [Link]
[116] G. Collett, N.J. Crammond, R.N. Swamy, J.H. Sharp, The role of carbon dioxide 9465(03)00155-0.
in the formation of thaumasite, Cem. Concr. Res. 34 (2004), [Link] [144] M. Romer, L. Holzer, M. Pfiffner, Swiss tunnel structures: Concrete damage by
10.1016/[Link].2004.02.024. formation of thaumasite, Cem. Concr. Compos. 25 (2003), [Link]
[117] S. Köhler, D. Heinz, L. Urbonas, Effect of ettringite on thaumasite formation, 10.1016/S0958-9465(03)00141-0.
Cem. Concr. Res. 36 (2006), [Link] [145] J.S. Šuput, A. Mladenovič, L. Černilogar, V. Olenšek, Deterioration of mortar
[118] W. Lukas, Betonzerstörung durch SO3-angriff unter bildung von thaumasit caused by the formation of thaumasite on the limestone cladding of some
und woodfordit, Cem. Concr. Res. 5 (1975) 503–517, [Link] Slovenian railway tunnels, Cem. Concr. Compos. 25 (2003), [Link]
0008-8846(75)90025-3. 10.1016/S0958-9465(03)00144-6.
[119] J. Bensted, Thaumasite––direct, woodfordite and other possible formation [146] A.J. Puppala, S. Saride, D. Dermatas, M. Al-Shamrani, V. Chikyala, Forensic
routes, Cem. Concr. Compos. 25 (2003) 873–877, [Link] investigations to evaluate sulfate-induced heave attack on a tunnel shotcrete
S0958-9465(03)00115-X. liner, J. Mater. Civ. Eng. 22 (2010), [Link]
[120] J. Bensted, Thaumasite - background and nature in deterioration of cements, 5533.0000087.
mortars and concretes, Cem. Concr. Compos. 21 (1999) 117–121, [Link] [147] N. Crammond, The occurrence of thaumasite in modern construction - a
org/10.1016/S0958-9465(97)00076-0. review, Cem. Concr. Compos. 24 (2002), [Link]
[121] M.M. Rahman, M.T. Bassuoni, Thaumasite sulfate attack on concrete: (01)00092-0.
Mechanisms, influential factors and mitigation, Constr. Build. Mater. 73 [148] F. Mittermayr, C. Bauer, D. Klammer, M.E. Böttcher, A. Leis, P. Escher, M.
(2014) 652–662, [Link] Dietzel, Concrete under sulphate attack: an isotope study on sulphur sources,
[122] T. Schmidt, B. Lothenbach, M. Romer, K. Scrivener, D. Rentsch, R. Figi, A Isotopes Environ. Health Stud. 48 (2012), [Link]
thermodynamic and experimental study of the conditions of thaumasite 10256016.2012.641964.
formation, Cem. Concr. Res. 38 (2008), [Link] [149] S.A. Hartshorn, J.H. Sharp, R.N. Swamy, The thaumasite form of sulfate attack
cemconres.2007.11.003. in Portland-limestone cement mortars stored in magnesium sulfate solution,
[123] F. Bellmann, J. Stark, Prevention of thaumasite formation in concrete exposed Cem. Concr. Compos. 24 (2002) 351–359, [Link]
to sulphate attack, Cem. Concr. Res. 37 (2007), [Link] 9465(01)00087-7.
cemconres.2007.04.007. [150] M.E. Gaze, N.J. Crammond, Formation of thaumasite in a cement:lime:sand
[124] B. Lothenbach, B. Bary, P. Le Bescop, T. Schmidt, N. Leterrier, Sulfate ingress in mortar exposed to cold magnesium and potassium sulfate solutions, Cem.
Portland cement, Cem. Concr. Res. 40 (2010) 1211–1225, [Link] Concr. Compos. 22 (2000) 209–222, [Link]
10.1016/[Link].2010.04.004. 00002-0.
[125] J. Wang, D. Niu, Y. Wang, B. Wang, Durability performance of brine-exposed [151] W. Lukas, H. Huber, Sulfate resistant shotcrete for road tunnels linning (in
shotcrete in salt lake environment, Constr. Build. Mater. 188 (2018) 520–536, German), Straßenforsch. Bundesministerium Für Baut. und Tech. Heft 192, 1982.
[Link] [152] C. Shi, D. Wang, A. Behnood, Review of thaumasite sulfate attack on cement
[126] D.A. St John, An unusual case of ground water sulphate attack on concrete, mortar and concrete, J. Mater. Civ. Eng. 24 (2012) 1450–1460, [Link]
Cem. Concr. Res. 12 (1982), [Link] 10.1061/(ASCE)MT.1943-5533.0000530.
[127] G.W. Scherer, Stress from crystallization of salt, Cem. Concr. Res. 34 (2004) [153] P.W. Brown, R.D. Hooton, B.A. Clark, The co-existence of thaumasite and
1613–1624, [Link] ettringite in concrete exposed to magnesium sulfate at room temperature and
[128] N. Thaulow, S. Sahu, Mechanism of concrete deterioration due to salt the influence of blast-furnace slag substitution on sulfate resistance, Cem.
crystallization, Mater. Charact. 53 (2004) 123–127, [Link] Concr. Compos. 25 (2003), [Link]
[Link].2004.08.013. [154] F. Mittermayr, M. Thumann, A. Baldermann, B. Lindlar, C. Stenger, H. Huber, E.
[129] C. Paglia, F. Wombacher, H. Böhni, The influence of alkali-free and alkaline Bauer, R. Scheutz and W. Kusterle, Repair of the Bosruck railroad tunnel:
shotcrete accelerators within cement systems: Influence of the temperature development of sprayed concrete with increased durability against
on the sulfate attack mechanisms and damage, Cem. Concr. Res. 33 (2003), thaumasite sulfate attack, in: W. Kusterle, (Ed), Spritzbet. Tagung, 2018;
[Link] Alpbach, Austria.
[130] C. Paglia, F. Wombacher, H. Böhni, M. Sommer, An evaluation of the sulfate [155] T.C. Powers, Freezing effects in concrete, Durab. Concr. (1975) 1–11.
resistance of cementitious material accelerated with alkali-free and alkaline [156] J.J. Beaudoin, C. MacInnis, Dimensional changes of hydrated Portland cement
admixtures: Laboratory vs. field, Cem. Concr. Res. 32, 2002, [Link] mortar due to slow cooling and warming, Durab. Concr. (1975).
10.1016/S0008-8846(01)00739-6 [157] S. Chatterji, Aspects of the freezing process in a porous material-water
[131] S.T. Lee, D.G. Kim, H.S. Jung, Sulfate attack of cement matrix containing system Part 1. Freezing and the properties of water and ice, Cem. Conc. Res.
inorganic alkali-free accelerator, KSCE J. Civ. Eng. 13 (2009), [Link] 29 (1999), [Link]
10.1007/s12205-009-0049-0. [158] S. Chatterji, Aspects of freezing process in porous material-water system Part
[132] M.T. Blanco-Varela, P.M. Carmona-Quiroga, I.F. Sáez Del Bosque, S. Martínez- 2. Freezing and properties of frozen porous materials, Cem. Concr. Res. 29
Ramírez, Role of organic admixtures on thaumasite precipitation, Cem. Concr. (1999).
Res. 42 (2012), [Link] [159] J.P. Kaufmann, Experimental identification of ice formation in small concrete
[133] J. Wang, D. Niu, R. Ma, Y. Zhang, Investigation of sulfate attack resistance of pores, Cem. Concr. Res. 34 (2004), [Link]
shotcrete under dry-wet cycles, J. Wuhan Univ. Technol. Mater. Sci. Ed. 31 01.022.
(2016), [Link] [160] O. Coussy, P.J.M. Monteiro, Poroelastic model for concrete exposed to freezing
[134] D.-T. Niu, Y.-D. Wang, R. Ma, J.-B. Wang, S.-H. Xu, Experiment study on the temperatures, Cem. Concr. Res. 38 (2008), [Link]
failure mechanism of dry-mix shotcrete under the combined actions of cemconres.2007.06.006.
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 491

[161] M. Eglinton, 7 – Resistance of Concrete to Destructive Agencies, in: Lea’s Chem. [189] F.P. Glasser, M. Tyrer, K. Quillin, D. Ross, J. Pedersen, K. Goldthorpe, D. Bennett
Cem. Concr., 1998, pp. 299–342. doi:10.1016/B978-075066256-7/50019-9. and M. Atkins, The chemistry of blended cements and backfills intended for
[162] G. Fagerlund, The critical degree of saturation method of assessing the freeze/ use in radioactive waste disposal, Chem. Blended Cem. Backfills Intend. Use
thaw resistance of concrete, Matériaux Constr. 10 (1977) 217–229, https:// Radioact. Waste Dispos., 1999.
[Link]/10.1007/BF02478693. [190] I. Galan, F.P. Glasser, Chloride in cement, Adv. Cem. Res. 27 (2015), https://
[163] T. Beck and K. Brusletto, Sprayed concrete used as avalanche securing of road [Link]/10.1680/adcr.13.00067.
in the open day during freezing conditions, Eighth Int. Symp. Sprayed Concr., [191] P. Hagelia, Sprayed concrete in aggresive subsea environment the Oslofjord
2018; Trondheim, Norway. test site, in: 6th Int. Symp. Sprayed Concr., 2011; Tromsø, Norway.
[164] M. Jolin, D. Beaupré, M. Pigeon, A. Lamontagne, Use of set accelerating [192] I. Galan, L. Perron, F.P. Glasser, Impact of chloride-rich environments on
admixtures in dry-mix shotcrete, J. Mater. Civ. Eng. 9 (1997). cement paste mineralogy, Cem. Concr. Res. 68 (2015), [Link]
[165] H.-G. Park, S.-K. Sung, C.-G. Park, J.-P. Won, Influence of a C12A7 mineral- 10.1016/[Link].2014.10.017.
based accelerator on the strength and durability of shotcrete, Cem. Concr. [193] J. Wang, D. Niu, Y. Zhang, Mechanical properties, permeability and durability
Res. 38 (2008), [Link] of accelerated shotcrete, Constr. Build. Mater. 95 (2015), [Link]
[166] J.-P. Won, U.-J. Hwang, S.-J. Lee, Enhanced long-term strength and durability 10.1016/[Link].2015.07.148.
of shotcrete with high-strength C12A7 mineral-based accelerator, Cem. [194] E. Nordström, Durability of Sprayed Concrete. Steel fibre corrosion in cracks,
Concr. Res. 76 (2015), [Link] Lulea University of Technology, Sweden, 2005.
[167] S. Lee, D. Kim, J. Ryu, S. Lee, J. Kim, H. Kim, M. Choi, An experimental study on [195] Y. Ouyang, X. Chen, C. Lu, J. Fang, Properties of steel fiber and macro-synthetic
the durability of high performance shotcrete for permanent tunnel support, fiber reinforced shotcrete, Dongnan Daxue Xuebao (Ziran Kexue Ban)/Journal
Tunn. Undergr. Sp. Technol. 21 (2006). Southeast Univ. (Natural Sci. Ed.) 40 (2010).
[168] Z. Hu, H. Ding, J. Lai, H. Wang, X. Wang, S. He, The durability of shotcrete in [196] Y.-W. Lin, A. Scott, L. Wotherspoon, J.M. Ingham, Durability properties of
cold region tunnel: a review, Constr. Build. Mater. 185 (2018) 670–683, sprayed engineered cementitious composite, ACI Mater. J. 110 (2013).
[Link] [197] A.G. Nitschke, R. Winterberg, Performance of macro synthetic fiber reinforced
[169] K.G. Holter, S. Smeplass, S. Jacobsen, Freeze–thaw resistance of sprayed tunnel linings, in: ITA-AITES World Tunn. Congr., 2016; San Francisco,
concrete in tunnel linings, Mater. Struct. Constr. 49 (2016), [Link] California.
10.1617/s11527-015-0705-4. [198] J.P. Kaufmann, Durability performance of fiber reinforced shotcrete in
[170] K.G. Holter, S. Geving, Moisture Transport Through Sprayed Concrete Tunnel aggressive environment., in: World Tunn. Congr., 2014; Foz do Iguaçu, Brasil.
Linings, Rock Mech. Rock Eng. 49 (2016), [Link] [199] S. Myren, P. Hagelia and P. Bjontegaard, The ban of polymer fibre in FRSC in
015-0730-1. Norwegian Road Tunnels, Eighth Int. Symp. Sprayed Concr., 2018;
[171] G. Mainali, S. Dineva, E. Nordlund, Experimental study on debonding of Trondheim, Norway.
shotcrete with acoustic emission during freezing and thawing cycle, Cold [200] L. Zhang, D. Morgan, S. Mindess, Comparative evaluation of transport
Reg. Sci. Technol. 111 (2015), [Link] properties of shotcrete compared to cast-in-place concrete, ACI Mater. J.
[Link].2014.11.014. 113 (2016), [Link]
[172] J. Chen, X. Deng, Y. Luo, L. He, Q. Liu, X. Qiao, Investigation of microstructural [201] P. Choi, K.-K. Yun, J.H. Yeon, Effects of mineral admixtures and steel fiber on
damage in shotcrete under a freeze-thaw environment, Constr. Build. Mater. rheology, strength, and chloride ion penetration resistance characteristics of
83 (2015), [Link] wet-mix shotcrete mixtures containing crushed aggregates, Constr. Build.
[173] A. Lamontagne, M. Pigeon, R. Pleau, D. Beaupré, Use of air-entraining Mater. 142 (2017), [Link]
admixtures in dry-mix shotcrete, ACI Mater. J. 93 (1996). [202] H. Li, D. Yan, G. Chen, S. Xu, J. Liu, Y. Hu, Porosity, pore size distribution and
[174] J. Chen, X. Zhao, Y. Luo, X. Deng, Q. Liu, Investigating freeze-proof durability chloride permeability of shotcrete modified with nano particles at early age,
of C25 shotcrete, Constr. Build. Mater. 61 (2014), [Link] J. Wuhan Univ. Technol. Mater. Sci. Ed. 31 (2016), [Link]
[Link].2014.02.077. s11595-016-1413-9.
[175] P. Choi, J.H. Yeon, K.-K. Yun, Air-void structure, strength, and permeability of [203] ASTM C1202-12 Standard Test Method for Electrical Indication of Concrete’s
wet-mix shotcrete before and after shotcreting operation: The influences of Ability to Resist Chloride Ion Penetration, West Conshohocken, Pennsylvania,
silica fume and air-entraining agent, Cem. Concr. Compos. 70 (2016), https:// 2012
[Link]/10.1016/[Link].2016.03.012. [204] NT Build 492, Concrete, Mortar and Cement-based Repair Materials: Chloride
[176] W. Lukas, W. Kusterle, Development of a stabilization procedure for freeze- Migration Coefficient from Non-Steady-State Migration Experiments, Espoo,
thaw damaged concrete (in German), Schriftenr. Straßenforschung Heft 366, 1999
1987. [205] JTG/T B07-01-2006 Specification for deterioration prevention of highway
[177] W. Kusterle, Large-scale shotcrete repair of concrete affected by frost and de- construction structures, Beijing, 2006.
icing salts (in German), in: W. Lukas and W. Kusterle, (Eds.), Instandsetzung [206] T. Wechner, Challenges of wet mix shotcrete application in an hyperbaric
Und Erhaltung von Baut. Aus Bet. Und Stahlbet.,1988; Innsbruck, Austria. atmosphare, using the example of the tunnel Karl-Friedrich-Strasse in
[178] L. Du, K.J. Folliard, Mechanisms of air entrainment in concrete, Cem. Concr. Karlsruhe (in German), in: W. Kusterle (Ed.), Spritzbet. Tagung, 2018;
Res. 35 (2005), [Link] Alpbach, Austria.
[179] M. Jolin, B. Bissonnette, A decade of shotcrete research at Laval University, in: [207] A. Bertsch, Changes in the properties of shotcrete subjected to compressed air
D.R. Morgan, H.W. Parker (Eds.), Shotcrete Undergr. Support X - Proc. Tenth (in German), University of Innsbruck, Austria, 1992.
Int. Conf. Shotcrete Undergr. Support, Whistler, British Columbia, Canada, [208] J. Schreyer, Adapting the sprayed shotcrete lining to water control by
2006. [Link] compressed air and to better working conditions (in German), Stuva
[180] C. Kan, Z. Zhang, G. Xiong, X. Yang, On the freezing-thawing mechanism of Forschungsberichte, Spritzbet. Unter Druckluft. 5/81, 1981.
macro-PP fiber shotcrete, Mod. Tunn. Technol. 52 (2015), [Link] [209] L.S. Dent Glasser, N. Kataoka, The chemistry of ‘‘alkali-aggregate” reaction,
10.13807/[Link].2015.06.023. Cem. Concr. Res. 11 (1981), [Link]
[181] D.R. Morgan, N. McAskill, G.G. Carette, V.M. Malhotra, Evaluation of [210] L.S. Dent Glasser, N. Kataoka, On the role of calcium in the alkali-aggregate
polypropylene fiber reinforced high-volume fly ash shotcrete, ACI Mater. J. reaction, Cem. Concr. Res. 12 (1982), [Link]
89 (1992). 90080-1.
[182] M.H. Zhang, J. Mirza, V.M. Malhotra, Mechanical properties and freezing and [211] J.W. Pan, Y.T. Feng, J.T. Wang, Q.C. Sun, C.H. Zhang, D.R.J. Owen, Modeling
thawing durability of polypropylene fiber-reinforced shotcrete incorporating of alkali-silica reaction in concrete: a review, Front. Struct. Civ. Eng. 6
silica fume and high volumes of fly ash, Cem. Concr. Aggregates. 21 (1999). (2012).
[183] A. Lamontagne, M. Pigeon, The influence of polypropylene fibers and [212] T. Katayama, The so-called alkali-carbonate reaction (ACR) - Its mineralogical
aggregate grading on the properties of dry-mix shotcrete, Cem. Concr. Res. and geochemical details, with special reference to ASR, Cem. Concr. Res. 40
25 (1995), [Link] (2010) 643–675, [Link]
[184] S.W. Tang, Y. Yao, C. Andrade, Z.J. Li, Recent durability studies on concrete [213] T. Katayama, How to identify carbonate rock reactions in concrete, Mater.
structure, Cem. Concr. Res. 78 (2015), [Link] Charact. 53 (2004) 85–104, [Link]
cemconres.2015.05.021. [214] J. Burdin, N. Monin, The management of excavated materials from the Lyon-
[185] GB, GB/T 50082-2009, Standard for test methods of long-term performance Turin rail link project | Materialbewirtschaftung am Projekt
and durability of ordinary concrete, Beijing, 2009. Eisenbahnverbindung Lyon-Turin, Geomech. Und Tunnelbau. 2 (2009),
[186] M.J. Setzer, R. Auberg, Freeze-thaw and deicing salt resistance of concrete [Link]
testing by the CDF method CDF resistance limit and evaluation of precision, [215] N. Monin, J. Burdin, L. Brino, J. Colas, L. Divet, T. Chaussadent, Preliminary
Mater. Struct. 28 (1995) 16–31, [Link] tests on concrete aggregates with high sulphate content for tunnel lining, in:
[187] M.J. Setzer, P. Heine, S. Kasparek, S. Palecki, R. Auberg, V. Feldrappe, E. Siebel, World Tunn. Congr., 2013; Geneva, Switzerland.
D.J. Janssen, R. Auberg, D.H. Bager, G. Gudmundsson, S. Jacobsen, T. Miura, H. [216] M. Petitat, K. von Allmen, J. Burdin, Automation of rock selection and
Mihashi, V. Penttala, P. Petersson, E.M. Schulson, J. Stark, L. Tang, P. Ursella, aggregate quality for reuse in tunnelling and industry, Geomech. Und
Test methods of frost resistance of concrete: CIF-Test: Capillary suction, Tunnelbau. 8 (2015), [Link]
internal damage and freeze thaw test - reference method and alternative [217] R. Galler, Development of resource-efficient tunnelling technologies - Results
methods A and B, Mater. Struct. Constr. 37 (2004) 743–753, [Link] of the European research project DRAGON, Geomech. Und Tunnelbau. 8
10.1617/14143. (2015) 302–309, [Link]
[188] M.J. Setzer, G. Fagerlund, D.J. Janssen, CDF test-test method for the freeze- [218] S. Ritter, H.H. Einstein, R. Galler, Planning the handling of tunnel excavation
thaw resistance of concrete-tests with sodium chloride solution (CDF), Mater. material - a process of decision making under uncertainty, Tunn. Undergr. Sp.
Struct. 29 (1996) 523–528. Technol. 33 (2013) 193–201, [Link]
492 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493

[219] A. Leemann, C. Thalmann, W. Studer, Alkali-aggregate reaction in Swiss [248] W. De Muynck, D. Debrouwer, N. De Belie, W. Verstraete, Bacterial carbonate
tunnels, Mater. Struct. Constr. 38 (2005), [Link] precipitation improves the durability of cementitious materials, Cem. Concr.
[220] J.-P. Won, B.-R. Choi, J.-W. Lee, Experimental and statistical analysis of the Res. 38 (2008), [Link]
alkali-silica reaction of accelerating admixtures in shotcrete, Constr. Build. [249] V. Achal, A. Mukherjee, M.S. Reddy, Microbial concrete: Way to enhance the
Mater. 30 (2012), [Link] durability of building structures, J. Mater. Civ. Eng. 23 (2011), [Link]
[221] I. Sims, P. Nixon, RILEM recommended test method AAR-1: Detection of 10.1061/(ASCE)MT.1943-5533.0000159.
potential aikali-reactivity of aggregates - Petrographic method, Mater. Struct. [250] F. Nosouhian, D. Mostofinejad, H. Hasheminejad, Concrete Durability
Constr. 36 (2003), [Link] improvement in a sulfate environment using bacteria, J. Mater. Civ. Eng. 28
[222] R.E. Oberholster, G. Davies, An accelerated method for testing the potential (2016), [Link]
alkali reactivity of siliceous aggregates, Cem. Concr. Res. 16 (1986) 181–189, [251] C. Gaylarde, M. Ribas Silva, T. Warscheid, Microbial impact on building
[Link] materials: an overview, Mater. Struct. Constr. 36 (2003), [Link]
[223] M.-A. Bérubé, B. Fournier, Canadian experience with testing for alkali- 10.1617/13867.
aggregate reactivity in concrete, Cem. Concr. Compos. 15 (1993) 27–47. [252] S.S. Bang, J.J. Lippert, U. Yerra, S. Mulukutla, V. Ramakrishnan, Microbial
[224] B. Fournier, M.-A. Bérubé, Alkali-aggregate reaction in concrete: A review of calcite, a bio-based smart nanomaterial in concrete remediation, Int. J. Smart
basic concepts and engineering implications, Can. J. Civ. Eng. 27 (2000) 167– Nano Mater. 1 (2010), [Link]
191. [253] H. Kalhori, R. Bagherpour, Application of carbonate precipitating bacteria for
[225] M. Thomas, R. Hooton, C. Rogers, Prevention of Damage Due to Alkali- improving properties and repairing cracks of shotcrete, Constr. Build. Mater.
Aggregate Reaction (AAR) in Concrete Construction-Canadian, Cem. Concr. 148 (2017), [Link]
Aggregates 19 (1997). [254] A. Bertron, Understanding interactions between cementitious materials and
[226] R.D. Hooton, C. Rogers, C.A. MacDonald, T. Ramlochan, Twenty-year field microorganisms: a key to sustainable and safe concrete structures in various
evaluation of alkali-silica reaction mitigation, ACI Mater. J. 110 (2013) 539– contexts, Mater. Struct. Constr. 47 (2014), [Link]
548. 014-0433-1.
[227] A. Ansell, Investigation of shrinkage cracking in shotcrete on tunnel drains, [255] N. De Belie, J. Monteny, A. Beeldens, E. Vincke, D. Van Gemert, W. Verstraete,
Tunn. Undergr. Sp. Technol. 25 (2010), [Link] Experimental research and prediction of the effect of chemical and biogenic
[Link].2010.04.006. sulfuric acid on different types of commercially produced concrete sewer
[228] C.I. Goodier, S.A. Austin, P.J. Robins, Low-volume wet-process sprayed pipes, Cem. Concr. Res. 34 (2004), [Link]
concrete: hardened properties, Mater. Struct. Constr. 41 (2008), [Link] cemconres.2004.02.015.
org/10.1617/s11527-007-9222-4. [256] C. Grengg, F. Mittermayr, A. Baldermann, M.E. Böttcher, A. Leis, G. Koraimann,
[229] L. Malmgren, E. Nordlund, S. Rolund, Adhesion strength and shrinkage of P. Grunert, M. Dietzel, Microbiologically induced concrete corrosion: a case
shotcrete, Tunn. Undergr. Sp. Technol. 20 (2005), [Link] study from a combined sewer network, Cem. Concr. Res. 77 (2015), https://
[Link].2004.05.002. [Link]/10.1016/[Link].2015.06.011.
[230] B. Lagerblad, L. Fjällberg, C. Vogt, Shrinkage and durability of shotcrete, in: E. [257] C. Grengg, F. Mittermayr, G. Koraimann, F. Konrad, M. Szabó, A. Demeny, M.
S. Bernard (Ed.), Shotcrete Elem. a Syst. - Proc. 3rd Int. Conf. Eng. Dev. Dietzel, The decisive role of acidophilic bacteria in concrete sewer networks:
Shotcrete, 2010; Queenstown, New Zealand. a new model for fast progressing microbial concrete corrosion, Cem. Concr.
[231] W. Lukas, W. Kusterle, Durability of Shotcrete made with Different Res. 101 (2017) 93–101, [Link]
Accelerators. The Influence of Water Glass on the Technological Parameters [258] S. Wei, Z. Jiang, H. Liu, D. Zhou, M. Sanchez-Silva, Microbiologically induced
of Shotcrete., in: Shotcrete Undergr. Support V, 1990; Uppsala, Sweden. deterioration of concrete - a review, Brazilian J. Microbiol. 44 (2013) 1001–
[232] C.K.Y. Leung, R. Lai, A.Y.F. Lee, Properties of wet-mixed fiber reinforced 1007, [Link]
shotcrete and fiber reinforced concrete with similar composition, Cem. Concr. [259] S. Karačić, P. Hagelia, F. Persson and B.-M. Wilén, Microbial attack on subsea
Res. 35 (2005), [Link] sprayed concrete, Int. RILEM Conf. Microorg. Mater. Interact., RILEM
[233] L.E. Bryne, A. Ansell, J. Holmgren, Investigation of restrained shrinkage Publications, 2016.
cracking in partially fixed shotcrete linings, Tunn. Undergr. Sp. Technol. 42 [260] C. Grengg, F. Mittermayr, N. Ukrainczyk, G. Koraimann, S. Kienesberger, M.
(2014), [Link] Dietzel, Advances in concrete materials for sewer systems affected by
[234] D.R. Morgan, R. Heere, C. Chan, J.K. Buffenbarger and R. Tomita, Evaluation of microbial induced concrete corrosion: a review, Water Res. 134 (2018) 341–
shrinkage-reducing admixtures in wet and dry-mix shotcretes, Proc. Int. 352, [Link]
Conf. Eng. Dev. Shotcrete, 2001; Hobart, Australia. [261] M. Hauki and T. Beck, Utilization of modern IoT systems in sprayed concrete
[235] S. Nagataki, H. Gomi, Expansive admixtures (mainly ettringite), Cem. Concr. processes, Eight Int. Symp. Sprayed Concr., 2018; Trondheim, Norway.
Compos. 20 (1998). [262] B. Lindlar, L. Oblak, D. Lootens and C. Stenger, From tunnel to laboratory:
[236] W. Huang, Q. Ma, Microstructure and strength characteristics analysis of scaling of shotcrete testing, 7th Int. Symp. Sprayed Concr., 2014 Sandefjord,
shrinkage-compensating shotcrete, in: Adv. Mater. Res. - Int. Conf. Adv. Eng. Norway, 282– 286.
Mater. Technol., 2011; Sanya, China. [263] R.P. Salvador, S.H.P. Cavalaro, M. Cano, A.D. Figueiredo, Influence of spraying
[237] C. Paglia, The influence of calciumsulfoaluminate as accelerating component on the early hydration of accelerated cement pastes, Cem. Concr. Res. 88
within cemntitious systems, Swiss Federal Institute of Technology, Zürich, (2016).
2000. [264] T. Sofis, The durability of shotcrete-Looking back at projects from 30 to 40
[238] K. Smith, T. Atkinson, Factors to consider in using PP fibres in concrete to years ago, Shotcrete, Am. Shocrete Assoc. (2015) 18–20.
provide explosive spalling resistance in the event of a fire, in: E.S. Bernard [265] S. Yoshida, F. Taguchi, S. Yamanaka, H. Sato, Applicability of shotcrete for
(Ed.), Shotcrete Elem. a Syst. - Proc 3rd Int. Conf. Eng. Dev. Shotcrete, 2010; NATM using blast furnace slag cement, Shotcrete Undergr. Support X, 2006.
Queenstown, New Zealand. Whistler, British Columbia, Canada.
[239] R. Jansson, Fire spalling of concrete - A historical overview, in: Proc. 3rd Int. [266] S. Yoshida, F. Taguchi, T. Nawa, H. Watanabe, Sulfuric acid resistance of
Work. Concr. Spalling Due to Fire Expo. - MATEC Web Conf., EDP Sciences, belite-based cement concrete mixed with GGBFS, 6th International
2013; Paris, France. Conference on Concrete under Severe Conditions, CONSEC’10, Merida,
[240] P. Bamonte, P.G. Gambarova, A. Nafarieh, High-temperature behavior of Yucatan, Mexico, 2010.
structural and non-structural shotcretes, Cem. Concr. Compos. 73 (2016), [267] EN 13057, Products and systems for the protection and repair of concrete
[Link] structures. Test methods. Determination of resistance of capillary absorption,
[241] P. Bamonte, P.G. Gambarova, Properties of concrete subjected to extreme 2002.
thermal conditions, J. Struct. Fire Eng. 5 (2014), [Link] [268] ISO 7031, Concrete hardened-Determination of the depth of penetration of
2040-2317.5.1.47. water under pressure, 1983.
[242] A. Bilodeau, V.K.R. Kodur, G.C. Hoff, Optimization of the type and amount of [269] ÖNORM B 3303, Test methods for concrete-National application of testing
polypropylene fibres for preventing the spalling of lightweight concrete standards for concrete and its source materials, Vienna, 2002.
subjected to hydrocarbon fire, Cem. Concr. Compos. 26 (2004) 163–174, [270] ASTM C1012, Standard Test Method for Length Change of Hydraulic-Cement
[Link] Mortars Exposed to a Sulfate Solution, 2018. West Conshohocken,
[243] P. Kalifa, G. Chéné, C. Gallé, High-temperature behaviour of HPC with Pennsylvania.
polypropylene fibres - from spalling to microstructure, Cem. Concr. Res. 31 [271] EN 107-1, Cement -Part 1: Composition, specifications and conformity
(2001) 1487–1499, [Link] criteria for common cements, 2011.
[244] M. Zeiml, D. Leithner, R. Lackner, H.A. Mang, How do polypropylene fibers [272] CEN/TS 12390-9, Testing hardened concrete - Part 9: Freeze-]thaw resistance
improve the spalling behavior of in-situ concrete?, Cem Concr. Res. 36 (2006) with de-icing salts -Scaling, 2016.
929–942, [Link] [273] ASTM C666, Standard Test Method for Resistance of Concrete to Rapid
[245] P.C. Tatnall, Shotcrete in fires: Effects of fibers on explosive spalling, Freezing and Thawing, 2015. West Conshohocken, Pennsylvania.
Shotcrete. (2002) 10–12. [274] ASTM C671, Standard Test Method for Critical Dilation of Concrete Specimens
[246] R. Kompen, How the use of fibres has developed in Norway, in: K. Berg, (Ed), Subjected to Freezing, 1994. West Conshohocken, Pennsylvania.
5th Int., Symp. Sprayed Concr., 2008; Lillehammer, Norway. [275] EN 480-11, Admixtures for concrete, mortar and grout. Test methods.
[247] W.A. Kusterle, Sprayable fire-protective layers in traffic tunnels, in: S.E. Determination of air void characteristics in hardened concrete, 2005.
Bernard (Ed.), Shotcrete Elem. a Syst. - Proc. 3rd Int. Conf. Eng. Dev. Shotcrete, [276] ASTM C672, Standard Test Method for Scaling Resistance of Concrete Surfaces
2010; Queenstown, New Zealand. Exposed to Deicing Chemicals, 2012. West Conshohocken, Pennsylvania.
 I. Galan et al. / Construction and Building Materials 202 (2019) 465–493 493

[277] SS 137244, Concrete Testing-Hardened Concrete-Scaling at Freezing, 2005. [282] ASTM C1260, Standard Test Method for Potential Alkali Reactivity of
Stockholm. Aggregates (Mortar-Bar Method), 2014. West Conshohocken, Pennsylvania.
[278] EN 1504-3, Products and systems for the protection and repair of concrete [283] ASTM C1567, Standard Test Method for Determining the Potential Alkali-
structures. Definitions, requirements, quality control and evaluation of Silica Reactivity of Combinations of Cementitious Materials and Aggregate
conformity. Structural and non-structural repair, 2005. (Accelerated Mortar-Bar Method), 2013. West Conshohocken, Pennsylvania.
[279] EN 13687-1, Products and systems for the protection and repair of concrete [284] CSA-A23.2-25A, Test Method for Detection of Alkali-Silica Reactive Aggregate
structures. Test methods. Determination of thermal compatibility. Freeze- by Accelerated Expansion of Mortar Bars, 2014. Mississauga, Ontario.
thaw cycling with de-icing salt immersion, 2002. [285] ASTM C1293, Standard Test Method for Determination of Length Change of
[280] CSA-A23.2-27A, Standard Practice to Identify Potential for Alkali-Reactivity of Concrete Due to Alkali-Silica Reaction, 2018. West Conshohocken, Pennsylvania.
Aggregates and Measures to Avoid Deleterious Expansion in Concrete, 2014. [286] CSA A23.2-14A, Potential Expansivity of Aggregates; Procedure for Length
Mississauga, Ontario. Change Due to Alkali-Aggregate Reaction in Concrete Prisms, 2014.
[281] ASTM C295, Standard Guide for Petrographic Examination of Aggregates for Mississauga, Ontario.
Concrete, 2018. West Conshohocken, Pennsylvania.