Mathematical errors are omnipresent in physics and engineering. In fact, errors are unavoidable when conducting experiments, modelling physical phenomena and when solving theoretical models either by hand or on digital computers. The ubiquity of errors in accuracy and precision makes their quantification and control fundamental. The standardised definitions of modelling accuracy and calculation precision are discussed ; and an analysis of calculation precision control is presented for a numerical implementation of a multiphysics model used for a radioactive waste storage application.