Resistance of concrete containing ggbs to the thaumasite form of sulfate attack

A long-term laboratory study has investigated how cement-type, aggregate-type and curing, affect the susceptibility of concrete to the thaumasite form of sulfate attack (TSA). The cements were Portland cement (PC), sulfate-resisting Portland cement (SRPC) and a combination of 70% ground granulated blastfurnace slag (ggbs) with 30% PC. These were combined with various carbonate aggregates or a non-carbonate control. Initial curing was either in water or in air. Concrete cubes were immersed in four strengths of sulfate solution at 5 and 20 °C. This paper reports the results after up to six years of immersion in sulfate solution.

Deterioration, consistent with TSA, was observed on many of the PC and SRPC concretes that had been made with carbonate aggregate and stored in sulfate solutions at 5 °C, with SRPC providing no better resistance to TSA than PC. Good quality concretes made with 70%ggbs/30%PC showed high resistance to TSA and the presence of carbonate in the mix substantially improved their general sulfate resistance. An initial air-cure, proved beneficial against both the conventional and thaumasite form of sulfate attack.     

About thaumasite formation in Portland-limestone cement pastes and mortars––effect of heat treatment at 95 °C and storage at 5 °C

Owing to the presence of finely divided calcite, mortars and concretes made with Portland-limestone cements are particularly susceptible to damaging thaumasite formation during sulfate attack at lower temperatures. This work reports the results of investigations on mortars made according to DIN/EN 196 and pastes (w/c ratio of 0.5) with CEM I 42,5 R, as well as with mixtures of cement with limestone filler. Some of the samples were heat-treated at 95 °C. The length changes and resonant frequencies of the samples were measured during long-term water-storage at 20 and 5 °C. There was no evidence from X-ray diffraction data of thaumasite formation in the samples. Only for pastes containing 30 wt.% limestone filler were small areas found by SEM and X-ray microanalysis whose chemical analysis matched thaumasite or a thaumasite–ettringite solid solution.     

Sulfate resistance of limestone cement concrete exposed to combined chloride and sulfate environment at low temperature

Concrete durability was investigated, taking under consideration the limestone content of the cement used, as well as the effect of chlorides on concrete’s deterioration due to the thaumasite form of sulfate attack. A normal Portland cement and two Portland limestone cements (15% and 35% w/w limestone content) were used for concrete preparation. The specimens were immersed in two corrosive solutions (chloride-sulfate; sulfate) and stored at 5 ± 1 °C. Visual inspection of the specimens, mass measurements and compressive strength tests took place for 24 months. Concretes containing limestone, as cement constituent and/or as aggregate, suffered from the thaumasite form of sulfate attack, which was accompanied by brucite and secondary gypsum formation. Limestone cement concretes exhibited higher deterioration degree compared to the concrete made without limestone cement. The disintegration was more severe and rapid, the higher the limestone content of the cement used. Chlorides inhibit sulfate attack on concrete, thus delaying and mitigating its deterioration.     

Thaumasite formed by sulfate attack on mortar with limestone filler

Early evidence of thaumasite formation in mortar with limestone filler exposed to sulfate containing tunnel water in Norway is reviewed. The problem is discussed in light of the new European cement standard allowing cements containing up to 35% limestone (e.g. CEM II/B-L) rendering them prone to detrimental sulfate attack.

Experiments are performed where mortars with 20% limestone or quartz filler, respectively, are stored in 5% sodium sulfate solution saturated with gypsum at 5 °C. Length change, flexural strength and compressive strength are measured periodically for a year. The microstructure of the mortars is inspected by scanning electron microscopy and energy dispersive analyser of X-rays documenting the formation of sulfate containing species including ettringite and thaumasite.     

Thaumasite formation in concrete and mortars containing fly ash

Due to recent reports on deterioration of concrete structures, the thaumasite form of sulfate attack has become a subject of study and close investigation. This paper investigates the formation of thaumasite in concrete and mortars containing fly ash. The results show that thaumasite formation can occur within 84 days of exposure to sulfate solutions. High volumes of fly ash can limit or promote thaumasite formation depending on the type of cement used. Thaumasite and ettringite were found among the deterioration products. However, the thaumasite formation in the specimen prepared from sulfate resisting Portland cement was not accompanied by deterioration, except by 50% fly ash addition. The mixtures of Portland limestone cement with 40% fly ash exhibited a very limited thaumasite formation while the mixtures with 50% had no thaumasite at all. It is concluded that thaumasite can also be formed in mixtures incorporating fly ash.     

The formation of thaumasite in a cement:lime:sand mortar exposed to cold magnesium and potassium sulfate solutions

This paper, which is presented in two parts, describes the performance of a laboratory-prepared masonry mortar after exposure to cold magnesium and potassium sulfate solutions. The objective of the study was to investigate the conditions under which the thaumasite form of sulfate attack can affect masonry mortars. This work was funded by the UK Department of the Environment, Transport and the Regions (DETR) and information contained in the following paper was included in the recently published DETR Report on the thaumasite form of sulfate attack (Department of the Environment, Transport and the Regions. The thaumasite form of sulfate attack: risks, diagnosis, remedial works and guidance on new construction. Report of the Thaumasite Expert Group, DETR, January 1999).In Part I of the investigation, mortar tablets (10 × 28 × 34 mm³) were prepared from a 1:1:5.5 cement:lime:sand, air entrained mortar to which powdered calcite had been added during mixing. Each mortar tablet was crushed before exposure in order to facilitate reaction and stored fully immersed in 200 ml of solution for several months. Two sets of storage conditions were used – one where the experiment was exposed to the atmosphere and the other where atmospheric exposure was prevented. The test material was sampled at intervals for analysis by X-ray diffraction to determine the nature of the products and the sequence of chemical events involved. Thaumasite was readily produced in both magnesium and potassium sulfate solution, following the prior formation of ettringite. It was not formed under conditions where ettringite was unstable, suggesting some involvement of the latter in the thaumasite forming process. It was also found that a rapid type of carbonation prevailed in potassium sulfate solution exposed to the atmosphere. This process, which has been called ‘alkali carbonation’, destroyed both ettringite and thaumasite.In Part II of this investigation, the same mortar tablets were used but this time the performance of the whole tablets of uncrushed mortar was tested in the same sulfate solutions so that the physical effects of sulfate attack on hardened mortars could be assessed. The test method used was similar in principle to the BRE mortar durability test which combines the cyclic administration of sulfate solution with intermediate drying to simulate the processes occurring in practice in brickwork mortar. X-ray diffraction analysis of the tablets was carried out when first signs of sulfate attack were observed and also after severe damage had occurred. These showed that early damage appeared to be mainly due to ettringite formation but both ettringite and thaumasite were involved at the severe damage stage. Some non-calcite containing mortars were also examined during this phase of the investigation and results have found them to be slightly less durable than their added-calcite counterparts, particularly in weak magnesium sulfate solution. This was attributed to the improved impermeability of added-calcite mortars rather than any inherent chemical resistance to sulfate attack.Even though Part I concentrated purely on the ‘chemical’ interactions between mortar and solution, the reaction products and sequences were found to be very similar to those discovered in Part II, where physical barriers to sulfate ingress had to be overcome prior to chemical attack. This provides confirmation that any masonry mortar can potentially deteriorate in the presence of excess sulfates providing the temperature is low, the mortar contains an available source of calcium carbonate, the brickwork is consistently wet and the pH of the reaction zone is maintained at 10.5 or above. Having said this, the extent of sulfate attack of brickwork in the field is small and should not become a major problem in practice, provided the current recommendations [Department of the Environment, Transport and the Regions. The thaumasite form of sulfate attack: risks, diagnosis, remedial works and guidance on new construction. Report of the Thaumasite Expert Group, DETR, January 1999] (especially the avoidance of using sulfate-bearing bricks in exposed situations) are adhered to.     

Investigations on the influence of cement type on thaumasite formation

A high sulfate resistance is required if cements are to be used in sulfate bearing waters and soils especially under conditions favouring thaumasite formation. A long period program of laboratory investigations was carried out on CEN cements to assess thaumasite and ettringite formation. The experimental concept involved mixing ground cement pastes with stoichiometrical components of gypsum, calcite and water. The specimens were stored at 6 °C whereby chemical worst case conditions for thaumasite formation were simulated. At time intervals XRD analysis was conducted. Apart from pure cements mixtures containing additives, pure C₃S pastes with and without Al₂O₃ addition were investigated. The results confirm that thaumasite formation can be accelerated by Al₂O₃ bearing components in cements. However, thaumasite formation is also possible without active participation of Al³⁺. The assessment of sulfate resistance of cements only from the chemical point of view apparently gives results which are contrary to the field experience.     

Performance of Portland limestone cements in mortar prisms immersed in sulfate solutions at 5 °C

The results of a test programme to investigate the sulfate resistance of mortars, immersed up to 12 months at 5 °C in magnesium sulfate and sodium sulfate solutions, is described. The mortars were prepared from four cements; a Portland cement, a sulfate-resisting Portland cement and two Portland limestone cements containing 15% by mass of an oolitic limestone and a carboniferous limestone. The mortar specimens were subject to BS 5328 Class 4A and 4B sulfate exposure conditions. These are the highest classes for concretes prepared using sulfate-resisting Portland cement (SRPC) before surface protection is required and are two and three classes higher than those recommended for concretes prepared using Portland cement (PC) and Portland limestone cement (PLC), respectively. Two free water-cement ratios were used, 0.5 and 0.75. Performance was monitored by visual assessment, expansion and changes in flexural and compressive strengths.At a free water-cement ratio of 0.75, the PC mortars and PLC mortars exhibited visually very severe attack with the former showing expansion and reductions in strength, and the latter mainly reductions in strength. At a free water-cement ratio of 0.50 both the PC mortars and PLC mortars showed slight/moderate to severe visual attack, the degree of deterioration appearing slightly greater in the PLC mortars, more especially those made with oolitic limestone. The PLC mortars also exhibited reductions in compressive failure load. The SRPC mortars exhibited little visual deterioration, no expansion, a small increase in flexural strength and no significant reductions in compressive strength. At a free water-cement ratio of 0.75 substantial amounts of thaumasite, together with ettringite was present in the surface layers of the deteriorated PLC mortars whilst ettringite was present in the surface layers of the deteriorated PC mortars. It is concluded that mortars made with a PC with a C₃A content of about 10% by mass were broadly similar in their vulnerability to sulfate attack at 5 °C as PLC mortars containing 15% limestone by mass, although the mode of attack was different.     

Use of mineral admixtures to improve the resistance of limestone cement concrete against thaumasite form of sulfate attack

In this work the effect of mineral admixtures on the thaumasite form of sulfate attack in limestone cement concrete is studied. Additionally, the effect of the type of sand (calcareous or siliceous) and the storage temperature is investigated. Limestone cement, containing 15% limestone, was used. Concrete specimens were prepared by replacing a part of cement with the studied minerals. The specimens were immersed in a 1.8% MgSO₄ solution and stored at 5 °C and 25 °C for 3 years. A well designed concrete made with limestone cement and fly ash, blastfurnace slag or metakaolin seems to have the ability to withstand thaumasite form of sulfate attack. The addition of natural pozzolana presented only a limited improvement of concrete’s sulfate resistance. The type of the sand and its cohesion with the cement paste has a remarkable effect on the performance of concrete at low temperature. Finally, no damage was observed in the specimens exposed to sulfate solution at 25 °C.     

An experimental study of combined acid and sulfate attack of concrete

There is disagreement about the role of sulfuric acid in the thaumasite form of sulfate attack (TSA) of concrete. Some researchers suggest that thaumasite is formed only at pH above 10.5, whereas others report that the primary cause of deterioration in the affected M5 bridge foundations was sulfuric acid attack followed by neutral TSA. The aim of this work is to reconcile these conflicting views by undertaking parallel studies of concrete exposed to aggressive acid and sulfate solutions and concrete/clay interface work using weathered Lower Lias clay.

Concrete specimens have been exposed to BRE Digest 363 sulfate class solutions and acidic and acidic-sulfate solutions at 4.5 ± 0.5 °C. Selected samples are being characterised at intervals up to 5 years. At this stage, results are reported for 5-month samples. Various binders including Portland cement, Portland–limestone cement, blastfurnace slag cement, pulverized-fuel ash cement and sulfate-resisting Portland cement at water/binder ratios (w/b) from 0.35 to 0.5 have been studied.

Initial visual observations and X-ray diffraction analyses have identified thaumasite in some of the systems after 5 months immersion in solution.

An overview of the ongoing parallel concrete/clay interaction work is also presented to contextualise the concrete work.     

Long term durability of Portland-limestone cement mortars exposed to magnesium sulfate attack

Mortar prisms made with Portland-limestone cement have been stored in air and in 1.8% magnesium sulfate solution at 5 °C and have been examined over a period of 5 years. This paper is primarily concerned with the results obtained at the end of this period. The limestone content in the samples varied from 0% to 35%, but the water to cement plus limestone powder ratio was kept constant. The status of the samples after storage for 5 years is reported based on visual examination and a thorough characterisation using X-ray diffraction, infra-red spectroscopy and scanning electron microscopy. The prisms stored in magnesium sulfate solution were all showing clear signs of deterioration, increasing in intensity with limestone content. The mortar prism with 5% limestone replacement was, however, seriously degraded in comparison with the ordinary Portland cement control prism, and it is shown that this was due to the thaumasite form of sulfate attack.     

Mechanism of thaumasite formation in concrete slabs on grade in Southern California

Thaumasite formation has been observed in residential concrete slabs on grade in Southern California. The concrete examined did not contain any carbonate bearing aggregates or fillers. Microstructural analyses showed a carbonated layer with calcite and gypsum at the bottom of the concrete. Above the carbonation layer, deposits of intermixed gypsum and thaumasite were observed. Further into the concrete towards the upper surface, deposits of thaumasite alone or in combination with ettringite were observed. Most of the thaumasite deposits were observed in air voids. SEM–EDS analysis showed deposits of ettringite, thaumasite and intermediate phases within the same air voids. The formation of thaumasite, ettringite and gypsum was caused by ingress of sulfate and carbonate ions from ground water. The presence of thaumasite, ettringite and intermediate phases in the same air void indicates that ettringite is first formed followed by thaumasite with a series of solid solutions. In this reaction process the pH of the local environment and the balance between sulfate, silicate and carbonate ions are important parameters.     

A thermodynamic model for predicting the stability of thaumasite

A model based on the phase rule has been used to predict the hydrate phase mineralogy and phase proportions from the chemical composition of hydrated Portland cement altered by sulfate attack. The eight-component system on which the model is based consists of CaO, SiO₂, Al₂O₃, Fe₂O₃, MgO, CaSO₄, CaCO₃ and H₂O. The phases included in the model are C–S–H, portlandite, ettringite, hydroxy-AFm, monosulfate, monocarbonate, calcite, gypsum, thaumasite, brucite and the pore solution. The model predicts, among other things, that thaumasite, which forms at low temperature, is unstable in the presence of AFm phases, and can only form in systems that would otherwise form gypsum at higher temperatures. The model has been tested experimentally on cement pastes containing 15 and 30 wt.% limestone dust stored at 5 °C, and which were either mixed with different amounts of gypsum and stored in water, or stored in solutions of different MgSO₄ concentrations. The fully hydrated pastes have been analysed by XRD and ²⁹Si CP/MAS NMR, whilst the remaining solution was analysed by ICP. Thaumasite is only found in regions where it has been predicted to form as a stable phase.     

Thaumasite sulfate attack in field and laboratory concretes: implications for specifications

The reported studies made on field elements affected by thaumasite sulfate attack are discussed, together with the reported laboratory performance of concrete and mortars immersed in sulfate solutions maintained at 5 °C. It is concluded that magnesium ions ingressing from groundwater or resulting from dedolomitization of an aggregate containing dolomite play a major role in increasing the risk of thaumasite attack and that attack can also arise from exposure to sulfuric acid bearing groundwaters, or sulfuric acid resulting from oxidation of pyrites within aggregates. Guidance is given on concrete qualities and materials to resist sulfate attack and thaumasite sulfate attack resulting from salts present in groundwater.     

Performance characteristics of concrete based on a ternary calcium sulfoaluminate–anhydrite–fly ash cement

This paper reports an assessment of the performance of concrete based on a calcium sulfoaluminate–anhydrite–fly ash cement combination. Concretes were prepared at three different w/c ratios and the properties were compared to those of Portland cement and blast-furnace cement concretes. The assessment involved determination of mechanical and durability properties. The results suggest that an advantageous synergistic effect between and ettringite and fly ash (Ioannou et al., 2014) was reflected in the concrete’s low water absorption rates, high sulfate resistance, and low chloride diffusion coefficients. However, carbonation depths, considering the dense ettringite-rich microstructure developed, were higher than those observed in Portland cement concretes at a given w/c ratio. It was concluded that the amount of alkali hydroxides present in the pore solution is as important factor as the w/c ratio when performance of this type of concrete is addressed.     

C, O and S isotopic signatures in concrete which have suffered thaumasite formation and limited thaumasite form of sulfate attack

Thaumasite formation and limited thaumasite sulfate attack has recently been discovered in sprayed concretes in contact with pyrrhotite-, pyrite- and calcite bearing Alum Shale in Oslo. In concretes, several types of calcite occur, including internal Popcorn calcite formed by replacement of both thaumasite and calcium silicate hydrate. In an attempt to throw further light on the origin of carbonates and sulfates involved, we have used the laser ablation probe to characterise these secondary minerals with respect to stable isotopes (C, O and S). Mitigation as well as repair may in several cases depend much on correct characterisation and location of the fluids provenance, and stable isotopic characterisation may be an appropriate tool to do so.

The preliminary results of this study indicate a complex open system with influence of fluids from several sources. There is a general difference in signatures between ordinary surface carbonation and internal carbonation associated with thaumasite. Calcite deposits within the Alum Shale/concrete contact zone show highly variable isotopic signatures reflecting a composite origin, probably significantly influenced by atmospheric CO₂. The sulfur isotopes in thaumasite appear to be much lighter than the Alum Shale constituents. This might possibly be explained by a contribution from atmospheric SO₂, or alternatively by sulfide oxidation in Alum Shale assisted by bacterial activity. A certain internal contribution from gypsum in cement clinker cannot be excluded.     

Theoretical analysis of the effect of weak sodium sulfate solutions on the durability of concrete

A theoretical analysis of the detrimental influence of weak sodium sulfate solutions (Na₂SO₄) on the durability of concrete is presented. It was conducted using a numerical model that takes into account the coupled transport of ions and liquid and the chemical equilibrium of solid phases within the (partially) saturated system. Numerous simulations were performed to investigate the influence of various parameters such as water/cement (w/c) ratio (0.45, 0.65 and 0.75), type of cement (CSA Type 10 and Type 50), sulfate concentration (0–30 mmol/l of SO₄) and the gradient in relative humidity across the material. All input data related to the properties of concrete were obtained by testing well-cured laboratory mixtures. Numerical results indicate that exposure to weak sulfate solutions can result in a significant reorganization of the microstructure of concrete. The penetration of sulfate ions into the material is not only at the origin of the precipitation of sulfate-bearing phases (such as ettringite and eventually gypsum) but also results in calcium hydroxide dissolution and C–S–H decalcification. Data also clearly emphasize the fact that w/c ratio remains the key parameter that controls the durability of concrete to sulfate attack.     

The laboratory investigation of concrete affected by TSA in the UK

Procedures that have been successfully employed by Geomaterials Research Services Ltd. in the determination of the distribution of the thaumasite form of sulfate attack (TSA) in concrete samples taken from bridge and other motorway structures throughout the UK are described. Electron microprobe analysis has been used to provide confirmation of the presence of thaumasite and to investigate the distribution of sulfate compounds including gypsum and ettringite in cement paste at the surfaces of concrete affected by sulfate attack. Electron microprobe analysis has the advantage that it is capable of detecting very small quantities of thaumasite. Electron microprobe analysis used in conjunction with petrographic analysis is regarded as the most effective tool for the diagnosis of TSA and other forms of sulfate attack.

The analysis of numerous cores taken from bridge foundations throughout the UK shows that the development of TSA is often accompanied by the formation of calcium carbonate in cement paste. High levels of chloride in the cement paste of TSA damaged concrete suggest the importance of run-off moisture as a contributory factor in the development of TSA.     

Sulfate resistance of cement-reduced eco-friendly concretes

Two newly developed cement-reduced eco-friendly concretes with high limestone powder content and low water/powder ratio were tested for sulfate resistance. Mortar samples with a paste composition of eco- as well as conventional concretes were immersed in 30 g l⁻¹ Na₂SO₄ and saturated Ca(OH)₂ reference solutions for 200 days at 8 °C. To evaluate the reaction mechanisms of progressing sulfate attack a combined approach of mechanical, mineralogical, and microstructural methods was applied.Gypsum and bassanite neo-formations related linearly to the expansion during sulfate exposure, except for one sample where ettringite co-precipitated. Thaumasite formation was not observed in spite of potentially favorable conditions. This is considered to be related to the evolution of the experimental solutions, kinetic effects, and the competing formation of CaCO₃ polymorphs triggered by the usage of superplasticizer. Both eco-friendly mixes exhibited a better sulfate resistance than their corresponding reference samples and are therefore suggested to be applicable in low sulfate-loaded environments according to DIN EN 206-1. Eco-friendly concrete based on CEM III/B performed superior against sulfate attack and is expected to withstand even severe sulfate exposure despite a much higher water/cement ratio than required by the standard.     

Microstructural changes in concretes with sulfate exposure

In prior papers the responses of concretes to 50,000 ppm MgSO₄ exposure depending on cement type, w/cm and the presence of slag were described. The present paper completes this analysis by examining the effects of immersion of concretes produced using slag blended cements, in solutions containing 50,000 ppm of sodium sulfate. The spatial evolution of microstructure associated with carbonation and sulfate attack show differences which can be related to the nature of the cation associated with the sulfate, the cement type, and the w/cm ratio.