In the Highway construction project development process, State Highway Agencies (SHA) prepare cost estimates for effective communication to stakeholders and for project cost control. Cost estimates prepared in the planning phase of project development typically in a time range of 10 to 20 years from project letting are characterized by a great deal of uncertainty due to low scope definition. SHAs typically include an amount as contingency in the project cost estimate to cover costs due to unidentified or unquantified risks during project development. However, most of the methods used by SHAs to apply contingency to projects lack consistency in definition and application. This leads to poor communication to stakeholders, project cost escalation and other project control issues due to inaccuracy of baseline cost estimates. This study developed a set of sliding scale contingencies for estimating contingency on highway projects taking into consideration the effect of major factors, such as project complexity that impacts contingency application. Expert opinion was sought through the use of the Delphi technique. Experimental techniques were not suitable for this study due to the exploratory nature of the problem and the lack of data to analyze using empirical methods. The Delphi method typically consists of a series of rounds called questionnaires. Twenty-three professionals with experience in risk assessment and cost estimating agreed to participate in the study. Email was the means of communication using an excel spreadsheet. The assessment was completed in three iterative rounds with controlled feedback to the participants on the panel at the end of each round. Sliding scale contingencies were developed for three levels of project complexity: noncomplex (minor), moderately complex, and most complex (major) projects. The sliding scale contingencies are presented as a final output of this study. This method of estimating contingency provides consistent rationale for estimating contingency. Risks are an inextricable part of the contingency estimating process. Estimators are encouraged to identify and document risks as justification for contingency values applied to a project.

Thesis (MEng (Civil Engineering))--University of Stellenbosch, 2006. / Most construction projects do not complete exactly as scheduled or exactly as priced. During the implementation of a project there is almost certainly some deviation from the original estimate. The implementation of a majority of projects has actually been shown to cost more and take longer than originally estimated. However, the duration and cost performance of a project’s implementation is measured against the initial estimate produced. Thus if a project is considered to have completed late or over budget then essentially the duration or cost estimated was insufficient. Due to the fact that estimates are produced in a present day environment for inherently unique projects that occur in uncertain future environments, the estimates produced will need to incorporate uncertainty to increase their likelihood of achievability. This study aims to derive methods to incorporate future uncertainty into project estimates. This uncertainty is incorporated, analysed and manipulated through the use of Probabilistic models and First Order Second Moment Reliability methods. The derived methods provide project management professionals with tools that enable them to design estimates that incorporate future uncertainty and are reliable to a specified degree. Further methods are then derived to probabilistically assess the commercial feasibility of a project in an uncertain future environment. These derived methods then provide project managers and decision makers with more reliable procedures and information which in turn should assist them in making correct, project orientated decisions and ultimately increase profit reliability and client satisfaction.