Operational risk is one of the core risks of every insurance company under the Solvency II framework and can be defined as the financial losses occurred due to incorrectly defined systems or processes; failures in IT system, human mistakes or other external processes. The research is performed in order to assess the capital to cover possible losses due to the occurrence of the operational risk sub-risks and nature of an operational risk. We have shown that operational risks can be modelled by skew t-copula and estimated tail dependence in each situation for modelling distributions with heavier tail area. The model is prepared on a non-life insurance company’s example and is based on the recorded data from loss database that encompasses historical information of five main operational sub-risks: legal, informational, organizational, human resources and expense risk.