Due to the complex structure, complicated operation process and extreme working environment, how to ensure the safe and efficient operation of the large-scale special equipment is always a big problem to be concerned. In this study, taking the large LNG storage tank as a research object, the methods and problems of the current risk analysis models of large-scale special equipment were investigated based on the proven numerical analysis technology combining with the big data model technology, Internet of things (IoT) and data sensing system. The method of creating the digital twin of a large LNG storage tank with process optimization, real-time monitoring and forecasting functions was analyzed. The framework of the digital twin was established according to its working process, and the key technologies were analyzed. It can provide references for the direction and method of establishing a relative perfect digital twin system of efficient optimization and risk analysis model for large special equipment.