The increasing complexity of power plants and the quest to decarbonize the energy sector require digital twin (DT) architecture to attain enhanced service and reduce maintenance costs. Nevertheless, DT applications in power plants are still nascent, with none focusing on gasification technology yet. This study intends to propose a robust DT architecture for a biomass gasification power plant (BGDT). An overview of the salient DT research in power plants was conducted to ascertain the current status and research gaps. DT classification was reviewed and applied to classify DTs available for power plants and their units. Thereafter, the requirements for the BGDT architecture were established and utilized to determine the main BGDT components. The components include a high-order scienceinformed dynamic model; a data-based model; actual data; preexecuted localized comprehensive simulations; and a system’s genome. The DT categorisation revealed a general gap in the mid-level of look-alike, and that most of the available power plant DTs lacked complete direct bi-directional data flow with the physical entity. The prototype dynamic model of the BGPP was developed using Aspen Plus software. Bagasse was deployed as the biomass in the model. With an air/fuel ratio of 2, a total of 32 MW power was achieved, and an energy analysis revealed energy savings of about 2%. This indicates that the developed model can serve as a twin to a physical plant or a basis for constructing a new one. Prospective studies include inculcating a CO2 removal facility into the model and developing other components of the BGDT.