Up to which dilution can one assume that 1 mg/l = 1 ppm?

Question

For aqueous solutions it is often assumed that 1 milligram per liter equals 1 ppm. Are there any guidelines up to which concentration this shortcut is admissible? Are there any formulas which calculate the error when using this shortcut?

Answer 1

"Parts per …" are dimensionless entities: to derive $\pu{ppm}$ from mass concentration $γ$ (e.g. $\pu{1 mg L-1})$ one uses density of water $\rho(\ce{H2O}) = \pu{1E6 mg L-1}$ for normalization:

$$\frac{γ}{ρ(\ce{H2O})} = \frac{\pu{1 mg L-1}}{\pu{1E6 mg L-1}} = \pu{1E-6} = \pu{1 ppm}$$

Once the solution is too concentrated, you obviously are not allowed to use $ρ(\ce{H2O})$ and have to use the real concentration of the solution. However, this occurs at the concentration levels way higher than those to be expressed with $\pu{ppm}$ anyway: that's why "parts per …" are so ubiquitously used in biochemistry. Higher molar masses, lower molar concentrations, primarily aqueous media: no need to account for the error stemming from the difference in the densities of pure water and the solution.

Note that "parts per something" units are deprecated and are not recommended for use, see e.g. Is 1 ppb equal to 1 μg/kg?.