Alkali–silica reaction (ASR)
The short of it.
This is a reaction between silica and potassium hydroxides, it occurs in an alkaline environment of pH 13 and higher. It forms a gel, and this gel expands as it absorbs water. The result is that the concrete cracks. It can take a long time before this is apparent. There are some spectacular examples of failure. (I insert the link when I find it again)
The Portland cement is the main source of the alkaline environment. Some Greywacke aggregates have been found particularly susceptible to ASR.
As FC is unlikely to contain Greywacke aggregates, there is little danger from that side, but there is foamed glass aggregate produced and these can react in the alkaline environment, unless they are AR-glass (Alkaline Resistant). However I like to be proven wrong, as this could be a good alternative.
The Scientific Explanation
Below is a more comprehensive explanation that I copied from:
Concrete Society Technical Report 30, 1999
Alkali–silica reaction (ASR) is potentially a very disruptive reaction within concrete. ASR involves the higher pH alkalis such as sodium and potassium hydroxides reacting with silica, usually within the aggregates, producing gel. This gel has a high capacity for absorbing water from the pore solution causing expansion and disruption of the concrete.
Some Greywacke aggregates have been found particularly susceptible to ASR. The main source of the alkalis is usually the Portland cement or external sources. Fly ash does contain some sodium and potassium alkalis but these are mainly held in the glassy structure and not readily available. Typically, only some 16–20 per cent of the total sodium and potassium alkalis in fly ash are water-soluble.
Many researchers have shown that fly ash is capable of preventing ASR. Oddly, the glass in fly ash is itself in a highly reactive fine form of silica. It has been found that the ratio of reactive alkalis to surface area of reactive silica is important in ASR. A pessimum ratio exists where the greatest expansion will occur. However, by adding more silica that is reactive plus the effect of the dilution of the alkalis mean that disruptive ASR cannot occur. The recommendations (BSI 5328 amendment 10365, 1999) within the UK require a minimum of 25 per cent BS 3892 Part 1 fly ash to prevent ASR. For coarser fly ashes, a minimum of 30 per cent fly ash may be required to ensure sufficient surface area to prevent ASR. Small quantities of fine fly ash with low-reactivity aggregates and sufficient alkalis may be more susceptible to ASR if the pessimum silica–alkali ratio is approached. Even when total alkalis within the concrete are as high as 5 kg/m³, fly ash has been found (Alasali and Mal-hotra, 1991) able to prevent ASR. The addition of fly ash reduces the pH of the pore solution to below 13 at which point ASR cannot occur. The use of low-alkali cements has a similar effect. However, the detailed mechanisms by which fly ash pre- vents ASR are complex and imperfectly understood.
The ACI Manual of Concrete Practice (1994) suggests few restrictions on the effectiveness of fly ashes. It states that ‘The use of adequate amounts of some fly ashes can reduce the amount of aggregate reaction’. It is later suggested that ashes only have to comply with ASTM C618, which permits a wide range of fly ashes.
Fournier and Malhotra (1997) investigated the ability of a range of fly ashes to prevent ASR. The activity index (AI) was found to affect the ASR performance. However, there was no correlation between fineness and AI. Nant-y-Moch, Dinas and Cwm Rheidol Dams are excellent examples of fly ash preventing ASR constructed about the same time using the same aggregates. The Nant-y-Moch and Cwm Rhei-dol dams used ‘run of station’ fly ash as part of the cement. These dams have performed well, in comparison to the Portland cement-only Dinas dam, which has some evidence of ASR cracking.