A study on dynamic cost risk early warning for EPC projects based on grey residual Markov prediction

Abstract

For engineering, procurement and construction(EPC) projects, the precise assessment and effective mitigation of cost overrun risks by managers are crucial for ensuring strict budget adherence and achieving the project's final profitability. However, existing cost risk management systems for EPC projects still suffer from issues such as delayed warnings and limited accuracy. To enhance the prediction and evaluation accuracy of cost overrun risks throughout the entire lifecycle of EPC projects, the advantages of Markov models and grey prediction models was integrated, applying a secondary Markov correction to the residuals of the grey prediction model. A dynamic early warning model for EPC project cost risks based on grey residual Markov prediction is constructed. Using actual case data from an EPC project, the proposed model's high accuracy and effectiveness in predicting cost overrun risks throughout the project lifecycle are validated. The case analysis result demonstrate that, compared to traditional grey prediction models, the grey residual markov prediction model offers significantly more accurate predictions of EPC project cost overruns, with the average residual error reduced from 4.27% to 1.74%. The findings not only enrich the theoretical framework for cost risk management in EPC projects but also provide practical support for managers in effectively assessing cost overrun risks at various stages of the project lifecycle.

Keywords

cost warning / grey residual markov prediction / EPC project