Evaluation of the Potential for Expansion and Cracking of Concrete Caused by the Alkali-Carbonate Reaction

Alkali-carbonate reactive rock was inadvertently used in the Cornwall and Ottawa areas of Ontario between 1978 and 1982. The excessive expansion and cracking caused by this reaction resulted in the need to replace concrete within three years of construction. This reaction of dolomitic limestone with the alkalies from cement is well known in the Midland-Kingston areas of Ontario. It had not been expected in the Cornwall and Ottawa areas. In an effort to prevent further occurrences of this reaction, 17 quarries in the Gulf River Formation were studied. Aggregate durability tests, chemical analysis, and concrete prism expansion tests were conducted on 26 bulk aggregate samples. Concrete prism expansion tests were conducted at alkali contents of 1.25 and 3.0% sodium oxide (Na2O) equivalent. Results showed that the current Canadian Standards Association specifications are inadequate for recognizing some alkali-carbonate reactive aggregates. This was due to the occurrence of delayed expansive aggregate. It is recommended that the concrete prism expansion test be conducted using cement with an alkali content of 1.25% Na2O. Those aggregates that exceed 0.025% expansion at one year should not be used in highway structures exposed to deicing salt. A rapid chemical screening test is proposed. The determination of the calcium oxide to magnesium oxide (CaO:MgO) ratio and alumina content or insoluble residue can be used to screen potentially alkali-carbonate expansive rocks from those that are nonexpansive.