Project Risk Aggregation for Portfolio Analysis - Framework Development for Vattenfall Services Nordic AB

Abstract

Organisations that manage multiple projects at the same time need to understand both the risks in each individual project and the overall risk exposure in the entire project portfolio. This thesis investigates how project risk information can be aggregated to support portfolio-level risk analysis. The research was conducted as a case study at Vattenfall Services Nordic AB (VSN) with the aim of developing a risk management framework. The study adopts an abductive research approach and combines a literature review, document analysis and 13 semi-structured interviews with tender engineers and project managers. In addition, a conceptual uncertainty analysis based on Monte Carlo simulation is developed to demonstrate how project risks can be analysed quantitatively and aggregated at portfolio level. The findings indicate that current risk identification and assessment practices are based on individual intuition and informal processes. Estimators use “embedded margins” directly within base cost estimates, which reduces transparency and obscures the actual risk exposure. Furthermore, inconsistent assessment methods make it impossible to aggregate risk data across the portfolio. To address these limitations, the thesis proposes a framework that introduces common principles for classification and assessment. Uncertainties are classified as uncertain known factors (Type A) or known uncertain factors (Type B), and deterministic single-value estimates are replaced by probabilistic analyses based on Monte Carlo simulations. The framework also supports continuous risk monitoring and feedback processes to strengthen organisational learning and knowledge transfer. The standardisation of project data enables the aggregation of individual project distributions into a combined portfolio-level distribution. This equips management with an overview of the total financial risk exposure. In the long term, the framework may support more informed decision-making and improved allocation of resources.