I am working on repairing some mildly deteriorated walls in a drinking-water treatment plant. There are no chlorides used in the treatment process. I would like to apply a 1 in.-thick shotcrete layer over the existing concrete utilizing a mix containing silica fume, which will achieve a compressive strength of 5000 psi at 28 days. I am having difficulty formulating a mix to meet those requirements that also has a water soluble chloride content of less than 0.10 % chloride ion concentration by mass of cement. I cannot get the chloride ion concentration below 0.15%. What adjustments can I make to get to my goal of 0.10% or less?

There are areas that have no problem getting values lower than the most stringent ACI requirement of 0.06% for prestressed concrete with no special adjustments. It would be prudent to test each of the proposed shotcrete constituents to determine their soluble chloride ion content. The most likely suspects are the aggregate and water sources. Typically portland cement and silica fume would contribute little, if any, detectable chloride ions. Assuming this would be a dry-process application, the only admixture other than the silica fume might be an air entraining agent, which would not provide any chloride ions. This leaves only the aggregates and water as the sources. At a minimum, the aggregates and water should be tested by a qualified laboratory for soluble chloride ion content. Alternate sources of aggregates and water may be required based on the laboratory results.