In the late 1930's parts of concrete highways in California were found to be cracking and buckling from, at that time, an unknown cause. The problem was investigated by Stanton, who was able to relate the deterioration to the type of cement used and the nature of the concrete aggregate. A good summary of the problem is given by Alderman (R1562), who states:

More potentially serious was the cracking of concrete in the Parker Dam, U.S.A. Reports on the deterioration were published by Blanks (R1565) and Meissner (R0187). Alderman states that:

Examination of drill cores stored in a damp environment showed the development of spots of a gel-like material which was largely sodium silicate. Since then, many examples of such phenomena have been encountered, and extensive investigations made to find solutions to the problem.

When it was realised that the disintegration of the concrete was due to cracking caused by expansive materials in the concrete, research projects were begun to identify the causes. The earliest work indicated that opaline silica was the main constituent of certain concrete aggregates in those cases where severe damage was caused to concrete structures.

It has also been recognised since then that there may be other constituents which can react with the cement and cause cracking; these include natural glasses as are found in some volcanic rocks, and possibly materials such as some clays and zeolites.

Early work in the United States by Stanton (R1554) indicated that the worst damage was caused by the presence of opaline silica in some aggregates. This was particularly serious when the cement used had a high alkali content (1% or more). Low alkali contents (< about 0.5%) rendered the aggregates substantially safe, although long term damage was still possible.

These observations were confirmed by work done in CSIR footnote in the mid 1940's by Alderman and others (R1562, R1563). They investigated the reaction of opal with Portland cement. The opal used was a material, probably opal-CT, from south eastern Victoria, Australia.

They found that, within a few days, damp spots would appear in the cement, small amounts of gel would form, and, later, cracks would develop. They found that this was due to the readiness with which opal would react with alkalies which were released from the cement when water was added.

A recent example of this phenomenon is illustrated in, where silica gel has formed, and cracked on partial drying. In this case, tiny opal or similar particles occurred in the aggregate used in a concrete floor. There did not appear to be enough to cause structural damage, but vinyl flooring laid on the concrete was damaged.

Portland cement consists of several phases; these include calcium silicates and aluminates which hydrate on the addition of water to form hydrates. The alkalis, both sodium and potassium, are present in the structures of these compounds, and are released on hydration. The alkalis form hydroxides which rapidly attack poorly crystalline materials such as opal and glass.

Nowadays, all aggregates from new areas are (or should be!) subjected to a thorough mineralogical examination to ensure that potentially reactive minerals are not present. In Australia, for example, the tough, hard siliceous rocks of central Australia have all the physical qualities needed for concrete aggregates, but in many cases detailed examination under the optical microscope reveals the presence of small amounts of opal in their structure. In addition, cement manufacturers have a much tighter quality control on the amount of alkali in their Portland cement.