The article features a critical observation on used methods for calculating the errors and a trial to improve it. The article describes the calculation method used in Sweden, county of Värmland and another method used by Ilvessalo and developed by Cajanus. The shortcomings of these models are discussed. An improved calculation is presented.