Most natural materials contain some chlorides and I agree with ACI 201 that no calcium chloride or admixtures that contain appreciable doses of calcium chloride should be added to concrete that will be exposed to moisture and chlorides in service, as in parking garages. But let's not go overboard and exclude the use of admixtures that contain small amounts of chlorides. In the past 30 years, admixtures have saved our industry billions of dollars in labor and materials. An overreaction could unnecessarily limit their use.Our company favors a balanced and objective approach when dealing with chloride limits in concrete. We do not favor an indiscriminate ban on the use of chloride-containing admixtures that would restrict the use of fly ash and pozzolans in cooler weather. Chloride limitations should be imposed on reinforced concrete in moist environments exposed to deicing salts and on prestressed concrete. However, ACI 201.2R-77 recommends no limit on chlorides for concretes that are above ground and dry. Furthermore, only the soluble chloride causes corrosion. A part of the chloride reacts with the calcium aluminate and ferrite phases to form insoluble chloride-containing compounds. Some of it also reacts with the silica-bearing phases, forming insoluble complexes. Bound or immobilized chloride cannot cause corrosion.

For many years, I was involved in the production of prestressed and other precast concrete products containing calcium chloride at levels of 1 and 2 percent of flake equivalent. Whenever good practice was observed, the concrete was trouble free. Most instances of failure involved complete disregard for cover over reinforcement and control of slump, and hence water-cement ratio. No cases are on record of problems with prestressed products where slump was controlled and water-cement ratio was low. The failure which caused the banning of calcium chloride back in the 1950's involved a prestress-steel winding on a pipeline, where the offending concrete was spray-applied. No doubt the prime cause of breakdown was incomplete cover over the reinforcement. Often it is convenient to be able to blame a chemical.