While runaway electron physics is a branch of tokamak research recently focusing on investigating, suppressing and mitigating it in the disruption phase, start-up runaways have been little interest after an early tokamak start-up study established. However, a recent ITER plasma initiation revision [1] indicated that ITER plasmas have a chance to produce the runaway electrons because of a low P plasma initiation condition constrained by a burn-through limit. ITER plasmas must avoid a runaways-dominant plasmas initiation, where runaway electrons carry most of plasma current, to prevent a harmful effect on the device by the energetic electrons. In fact, JET start-up runaway study reported that there is a case where the runaway current dominated the thermal current, which prohibited the thermal population from burning [2]. The runaway dominancy seems to arise from electrons density and electric field dynamics at an early stage. Understanding of their formation and energy limit is necessary in order to predict how much they do the harm on the device. In this work, we only tackle the modelling of their formation process. We conducted a numerical analysis on a formation of runaway electrons at KSTAR Ohmic breakdown start up employing the gyro-averaged Fokker Planck solver [3] and the DYON [4]. An evaluation of a reduced screening effect on electron-ion elastic and electron-electron inelastic collision has been typically concentrating on high Z impurity because the high Z impurities like Ne and Ar were used for runaway mitigation [5-7]. However, a species is different at plasma initiation, i.g. C for KSTAR and Be for ITER, and thus low Z impurity’s effect on deflection and slowing-down frequency should be revisited. We extended an investigation to relevant low-Z elements. Before equilibrium reliably reconstructed, the plasma position cannot be located and thus pushing it inward is adopted to a position control strategy. In this process, impurities from the wall can penetrate into the plasma, which should affect electron density, electric field and impurity contents. The DYON code can capture this picture setting the sputtering coefficient as a function of time. The well-known runaway generation mechanisms are the Dreicer mechanism, the avalanche mechanism and the hot-tail mechanism. The last one omitted in 0d modelling but the others cannot describe a time transient effect. Moreover, the formulas for 0d modelling were developed assuming fully-ionized plasmas and thus they are not applicable at an early stage of tokamak start-up. On the other hands, the gyro-averaged Fokker Planck solver can be applicable to weakly ionized plasmas and consider all mechanisms with neutral and impurity only if the maxwellian background assumption is valid. With the DYON and the gyro-averaged Fokker Planck solver, we modelled the formation of start-up runaways at KSTAR Ohmic plasma. We compared the Dreicer flow rate and the runaway current between the analytic expectation and the numerical computation. This research was supported by R&D Program of “R&D on Key Technology of ITER Components – Study on start-up supra-thermals/runaways for ITER plasmas operation (code No. 0644-20210066)” through the Korea Institute of Fusion Energy (KFE) funded by the Government funds, Republic of Korea. [1] P. C. de Vries and Y. Gribov 2019 Nucl. Fusion 59 096043 [2] P. C. de Vries et al 2020 Plasma Phys. Control. Fusion 62 125014 [3] Pavel Aleynikov and Boris N. Breizman 2017 Nucl. Fusion 57 046009 [4] Hyun-Tae Kim et al 2012 Nucl. Fusion 52 103016 [5] J. R. Martín-Solís et al 2017 Nucl. Fusion 57 066025 [6] Boris N. Breizman et al 2019 Nucl. Fusion 59 083001 [7] L. Hesslow et al, Phys. Rev. 118, 255001 (2017)