Despite the multidisciplinary networks involved in refurbishment of complex building projects, the lack of BIM adoption signifies lack of real BIM benefits towards acceptance of BIM within Architecture, Engineering and Construction (AEC).

To this end, this study empirically examines the potential real benefits between traditional network and BIM network for a real-time refurbishment case study project, through agentbased simulation modelling. A social network analysis theory is adapted to model the project interaction networks and a BIM prototype network. An assessment of the main stakeholders for BIM perception is carried out. We offered three prototype interaction networks for comparison of real BIM benefit.

An agent-based Bayesian network model is used to simulate the propagation of design error within the project networks. The result of the analysis show that BIM project diffuses error efficiently, while stakeholders recovers faster and nearly at the same time than traditional network. The optimised network shows better performance to the traditional network, when there is early involvement of subcontractors. The main contribution of this study is providing a novel approach to compare real benefits for traditional method to BIM method for refurbishment project and to provide avenue for project stakeholders to optimise their interaction through adoption of BIM.