Improving Cost Predictability in Dutch Construction Projects

Abstract

Cost overruns and cost deviations remain a persistent challenge in Dutch construction and infrastructure projects. While such deviations are usually visible in financial reporting once they occur during execution, their underlying causes are often not systematically traceable. This limits the ability of project organisations to learn from cost deviations and to improve the cost predictability of future projects. This thesis addresses this problem through the main research question: "How can cost predictability be improved in Dutch construction projects?"

The research starts from the premise that cost predictability depends not only on more accurate estimating or forecasting, but also on how cost information is structured, transferred, and used throughout the project. As cost information is repeatedly restructured between tendering, work budgeting, execution, reporting, and evaluation, the original assumptions behind it can become difficult to recognise or lost altogether; leaving deviations financially visible but no longer causally traceable.

The study was conducted in three phases. First, a literature review examined the relationship between cost structuring, cost traceability, project learning, and cost predictability, establishing that cost traceability is a necessary condition for double-loop learning: organisations can only learn structurally from deviations if they understand not just that a deviation occurred, but why. Five structuring approaches, Work Breakdown Structure, Critical Path Method, Earned Value Management, Last Planner System, and Advanced Work Packaging, were analysed for their respective contributions to structuring scope, time, cost, progress, and execution readiness.

Second, the research empirically investigated current practice within Dura Vermeer through organisational interviews and three project cases: Highway A16 De Groene Boog Rotterdam, Dike Reinforcement Tiel-Waardenburg, and Substation Vierverlaten. The cross-case analysis found that the limitation is not a lack of cost information but a lack of structural continuity between them. Cost structures change as projects move through phases, weakening the link between original assumptions, execution conditions, and final outcomes. Indirect costs proved especially difficult to trace, and cost and physical progress were often insufficiently integrated.

Based on these findings, a Construction Project Cost Traceability Framework was developed, using the Construction Work Package as a stable linking unit connecting tender assumptions, work budgets, physical progress, deviations, change mechanisms, indirect cost drivers, and learning outputs. Rather than prescribing full implementation of any single method, the framework combines selected principles: Advanced Work Packaging provides the lifecycle-oriented package structure, Last Planner System supports execution readiness and feedback, and Earned Value Management supports cost-progress measurement at package level.

The framework was validated through expert interviews with experienced Dura Vermeer practitioners, who confirmed the problem diagnosis and considered the framework feasible under conditions such as disciplined cost-data input, aligned systems across phases, clear ownership of cross-project learning, and incremental, change-management-supported introduction.

The main conclusion is that cost predictability in Dutch infrastructure projects can be improved by strengthening causal cost traceability across project phases, keeping cost information connected to the assumptions, scope elements, and decisions from which deviations originate, enabling both better in-project explanation and more systematic organisational learning.