[en] A simplified procedure, using circular disk models with homogeneous electric conductivity as representations for different body parts, has been proposed recently by product standard IEC 62822-3 for the assessment of magnetic field exposure in proximity to current-carrying conductors of welding equipment. Based on such simplified models, worst case coupling coefficients CCE_i(I), i.e. maximum induced electric field strength, normalised for current and frequency, for body parts at different distances d to straight single and double wire arrangements, as well as rectangular loop-shaped current paths are tabulated in the standard. In this work we compared CCE_i(I) values obtained by numerical computations with detailed anatomical models of the hand/forearm with the corresponding values given in IEC 62822-3 for current-carrying single wire conductors along the forearm at distances d = 30, 50 and 100 mm, respectively. Our results clearly indicated that the CCE_i(I) given in the standard may substantially underestimate the actual exposure. Using average values for tissue conductivities the observed extent of underestimation was up to 8.9 dB (factor 2.79) and may be even higher for worst case combinations of tissue conductivities. The reasons for this substantial underestimation are the oversimplified geometry, i.e. the circular disk does not reflect anatomical constrictions of the induction area present in realistic hand/forearm geometries, as well as the missing conductivity contrast between different tissues in the homogeneous disk models. Results of exposure assessment and corresponding minimum distances to components of welding equipment obtained by the simplified disk model approach suggested by IEC 62822-3 should therefore be considered with caution. (paper)