In the manufacture of soda-lime-silica glass a number of minute inclusions are formed. There are at least fifty identified chemical types. These are generally simply appearance imperfections that are of little consequence except in cases where they are large enough to be easily seen. Allowable limits are listed in ATM C 1036-85.

In cases where nickel-rich contaminants, such as stainless steel, are in the glass batch materials, they may combine with sulphur in the batch or furnace fuels to form nickel sulphide inclusions. Extremely minute powders or grains of nickel-rich materials will create these inclusions. They are impossible to totally eliminate and all glass has some present.

When glass is heat-treated the nickel sulphide inclusions are modified into a form that grows or transforms with time and temperature. If the inclusions are near the thickness centre of the glass, the transformation of the inclusions may create sufficient stresses in the glass to cause spontaneous breakage. This is more likely to occur in fully-tempered glass than heat-strengthened glass.

The nickel sulphide inclusions are actually a chemical group having varying ratios of nickel to sulphur. The effect of this range is that some of the inclusions transform much more rapidly than others for the same combination of time and temperature.

To reduce the possibility of breakage from nickel sulphide inclusions in the field, accelerated exposures at high temperatures are sometimes conducted. The inclusions that transform at a rapid rate may fail in the test. The accelerated exposure or “heat soak” test may then reduce the likelihood of field breakage in some cases. Those inclusion that transform at a slow rate, however, may be transformed to the point of causing imminent failure. This glass would be likely to fail soon after installation. In the latter case the incidence of breakage in the field may be increased from the effects of the “heat soak” test. Since the manufacturer has no means of identifying the nickel to sulphur ratio in inclusions that may be present, the “heat soak” test can not be relied upon to result in reduced field breakage of fully-tempered glass.

Available data suggest that “heat soak” tests for a time period of five to eight times those presently used are necessary to provide assurance against field breakage from nickel sulphide inclusions. Because of excessive cost this extended test is not conducted by any manufacturer on a routine basis.