Success attained in work with synthetic recirculated water with a low ME of calcium carbonate and containing colloidal calcium carbonate has suggested that a practical field method might be devised for developing permanent protective coatings in relatively short periods of time. In order to explore this possibility, studies were begun with a continuous flow of Michigan State University tap water. In a series of tests, the tap water flow was divided to feed sodium hydroxide and sodium carbonate to a bypassed portion. The high, 10.5 pH level in the bypassed flow resulted in the development of colloidal calcium carbonate in that line. Data and qualitative observations reported here show that a useful approach to the problem of the formation of protective calcium carbonate coatings in iron water mains has been developed. The experimental data lead to these conclusions: Michigan State University tap water develops hard, dense, calcite coatings on iron nipples within a few hours at velocities of 8.0‐13.5 fps and a momentary calcium carbonate excess of about 5 ppm. Momentary excesses as high as 60 ppm, with a flow velocity of 1.5 fps, develop similar coatings, an uncontrolled momentary excess of more than 30 ppm results in heavy deposits of loose calcite, which vary from fine to coarse grain; the addition of O.5‐1.5 ppm metaphosphate controls calcium carbonate precipitation and permits the use of momentary‐excess values as high as 100 ppm without encrustation of feed lines or piping. With a velocity of 3.0 fps, thin, translucent calcite coatings have been developed which have useful anticorrosion value; calcite coatings have been obtained with metaphosphate concentrations of 2.2 ppm and calcium levels as high as 1,500 ppm (as CaCO3); however, these coatings are not superior to those obtained at lower metaphosphate and calcium levels; the DFI and ME values have proved to be more useful for predicting carbonate deposition than Langelier's or Ryznar's indexes; microscopic examination of the coatings developed on sandblasted metal has shown imperfections and rusting in the translucent calcite coatings, it is believed that imperfections are at points of deep penetration by high‐velocity sand grains during sandblasting; and, better coatings were obtained in 2‐hr experiments at 3.0 fps than at 1.5 fps. In general, better coatings have resulted at high velocity than at low velocity, probably because of the decreased thickness of the laminar boundary film, more rapid mixing rate, and increased velocity distribution.