Icing is a strong limitation for the performance of wind turbines in cold climates and the prediction of the performance loss due to ice accretion is essential for designing effective ice mitigation measures. This paper presents a numerical approach capable of simulating the ice accretion transient phenomenon and its effects on wind turbine performance. This approach is applied to the NREL 5 MW reference wind turbine to predict (i) its performance during and after an icing event lasting for 8 h and (ii) the potential improvement in energy harnessing due to different operational strategies. The results show that by decreasing the turbine rotational speed and accepting a slight energy conversion decrease during the icing event, the performance can improve up to 6% when full operation is restored compared to the baseline operational strategy. Whereas, sustaining the rotational speed during the icing event can generate a 3% of performance loss afterwards compared to the same baseline. The developed workflow can be used for optimising performance of wind turbines by accounting for environmental conditions, the duration of the icing event, and performance after the icing event itself, thus constituting a valuable tool to maximise profitability of wind turbines in cold climates. (C) 2017 The Authors. Published by Elsevier Ltd.