Three tenths of the square root of FA.
Taking Melbourne water as an example of a good water supply, typically the tap water is around 100 uS/cm: my tap this morning read 107.
If this were all sodium chloride* that would be about 50 ppm NaCl, that's 0.005%. Your refractometer would read that as about 0.006** Brix. This is within the error band of even a lab grade refractometer with a monochromatic light source.
Taking Adelaide as an example of a bad water supply, the conductivity of the water can be around 400 uS/cm, so multiply the above figures by four. The error is still only about 0.025 Brix.
*which it isn't but the other likely components are in the same order of magnitude for refractive index vs concentration.
** Because the relationship between concentration and refractive index changes with the solute. At the same concentration, sucrose will read about 20% higher than sodium chloride on a refactometer.