Analogical reasoning for risk assessment and cost overrun estimation in construction projects


Thesis Type: Postgraduate

Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Civil Engineering, Turkey

Approval Date: 2010

Student: ONUR ÇELENLİGİL

Co-Supervisor: MUSTAFA TALAT BİRGÖNÜL, İREM DİKMEN TOKER

Abstract:

Project cost increase is the main concern in international construction projects which usually results in disputes and conflicts among the project participants. The aim of this thesis is to construct a database that represents risk event history regarding international construction projects and construct a cost overrun prediction model. It is hypothesized that magnitudes of project related, company related and country related risk factors can be predicted by assessing the level of vulnerability by analogical reasoning with previous projects. The vulnerability and risk factors can further be used to predict cost overrun in the bid preparation stage of international construction projects. Thus, prediction models that link vulnerability with risk factors and cost are constructed by using a dataset of 166 international construction projects, which consists of 66 real and 100 hypothetical cases. Case-based reasoning (CBR) technique is used to construct the prediction models. After testing the performance of various CBR models using different weight generation and retrieval methods, error rate of +/- 7.15 % cost increase is achieved. The utilization of CBR models in the prediction of potential risk sources and cost overrun is demonstrated by a real case study. Finally, the benefits and pitfalls of using analogical reasoning for risk and cost overrun assessment of construction projects are discussed.